diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13333 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 496.73202614379085,
+  "eval_steps": 500,
+  "global_step": 19000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.26143790849673204,
+      "grad_norm": 0.7138202786445618,
+      "learning_rate": 0.00019999986330190926,
+      "loss": 2.463,
+      "step": 10
+    },
+    {
+      "epoch": 0.5228758169934641,
+      "grad_norm": 0.790814220905304,
+      "learning_rate": 0.00019999945320801072,
+      "loss": 2.1963,
+      "step": 20
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 0.8431825041770935,
+      "learning_rate": 0.00019999876971942557,
+      "loss": 1.7823,
+      "step": 30
+    },
+    {
+      "epoch": 1.0457516339869282,
+      "grad_norm": 0.8540534377098083,
+      "learning_rate": 0.0001999978128380225,
+      "loss": 1.4428,
+      "step": 40
+    },
+    {
+      "epoch": 1.3071895424836601,
+      "grad_norm": 0.8526931405067444,
+      "learning_rate": 0.00019999658256641747,
+      "loss": 1.1373,
+      "step": 50
+    },
+    {
+      "epoch": 1.5686274509803921,
+      "grad_norm": 1.1626569032669067,
+      "learning_rate": 0.00019999507890797408,
+      "loss": 0.8482,
+      "step": 60
+    },
+    {
+      "epoch": 1.8300653594771243,
+      "grad_norm": 1.191985845565796,
+      "learning_rate": 0.0001999933018668033,
+      "loss": 0.6793,
+      "step": 70
+    },
+    {
+      "epoch": 2.0915032679738563,
+      "grad_norm": 0.8180794715881348,
+      "learning_rate": 0.0001999912514477634,
+      "loss": 0.5741,
+      "step": 80
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 0.7215670347213745,
+      "learning_rate": 0.00019998892765646026,
+      "loss": 0.424,
+      "step": 90
+    },
+    {
+      "epoch": 2.6143790849673203,
+      "grad_norm": 0.9890243411064148,
+      "learning_rate": 0.0001999863304992469,
+      "loss": 0.4335,
+      "step": 100
+    },
+    {
+      "epoch": 2.8758169934640523,
+      "grad_norm": 0.8641286492347717,
+      "learning_rate": 0.00019998345998322397,
+      "loss": 0.4301,
+      "step": 110
+    },
+    {
+      "epoch": 3.1372549019607843,
+      "grad_norm": 0.8116602301597595,
+      "learning_rate": 0.0001999803161162393,
+      "loss": 0.3547,
+      "step": 120
+    },
+    {
+      "epoch": 3.3986928104575163,
+      "grad_norm": 0.9898081421852112,
+      "learning_rate": 0.0001999768989068881,
+      "loss": 0.3373,
+      "step": 130
+    },
+    {
+      "epoch": 3.6601307189542482,
+      "grad_norm": 0.762677788734436,
+      "learning_rate": 0.0001999732083645129,
+      "loss": 0.3405,
+      "step": 140
+    },
+    {
+      "epoch": 3.9215686274509802,
+      "grad_norm": 0.9873971343040466,
+      "learning_rate": 0.0001999692444992035,
+      "loss": 0.3842,
+      "step": 150
+    },
+    {
+      "epoch": 4.183006535947713,
+      "grad_norm": 0.9828356504440308,
+      "learning_rate": 0.00019996500732179695,
+      "loss": 0.313,
+      "step": 160
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 0.8973499536514282,
+      "learning_rate": 0.0001999604968438775,
+      "loss": 0.3191,
+      "step": 170
+    },
+    {
+      "epoch": 4.705882352941177,
+      "grad_norm": 0.9795909523963928,
+      "learning_rate": 0.0001999557130777767,
+      "loss": 0.3101,
+      "step": 180
+    },
+    {
+      "epoch": 4.967320261437909,
+      "grad_norm": 1.2029128074645996,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 0.3403,
+      "step": 190
+    },
+    {
+      "epoch": 5.228758169934641,
+      "grad_norm": 1.0984058380126953,
+      "learning_rate": 0.00019994532573409262,
+      "loss": 0.2631,
+      "step": 200
+    },
+    {
+      "epoch": 5.490196078431373,
+      "grad_norm": 0.8392295241355896,
+      "learning_rate": 0.0001999397221849079,
+      "loss": 0.2942,
+      "step": 210
+    },
+    {
+      "epoch": 5.751633986928105,
+      "grad_norm": 0.8194178938865662,
+      "learning_rate": 0.00019993384540433894,
+      "loss": 0.3027,
+      "step": 220
+    },
+    {
+      "epoch": 6.0130718954248366,
+      "grad_norm": 1.1430736780166626,
+      "learning_rate": 0.00019992769540845258,
+      "loss": 0.312,
+      "step": 230
+    },
+    {
+      "epoch": 6.2745098039215685,
+      "grad_norm": 1.0457488298416138,
+      "learning_rate": 0.00019992127221406275,
+      "loss": 0.2631,
+      "step": 240
+    },
+    {
+      "epoch": 6.5359477124183005,
+      "grad_norm": 0.7075788974761963,
+      "learning_rate": 0.0001999145758387301,
+      "loss": 0.2814,
+      "step": 250
+    },
+    {
+      "epoch": 6.7973856209150325,
+      "grad_norm": 0.8417288064956665,
+      "learning_rate": 0.00019990760630076237,
+      "loss": 0.2945,
+      "step": 260
+    },
+    {
+      "epoch": 7.0588235294117645,
+      "grad_norm": 1.1337305307388306,
+      "learning_rate": 0.000199900363619214,
+      "loss": 0.288,
+      "step": 270
+    },
+    {
+      "epoch": 7.3202614379084965,
+      "grad_norm": 0.9480029940605164,
+      "learning_rate": 0.00019989284781388617,
+      "loss": 0.2696,
+      "step": 280
+    },
+    {
+      "epoch": 7.5816993464052285,
+      "grad_norm": 0.9333047270774841,
+      "learning_rate": 0.0001998850589053268,
+      "loss": 0.2647,
+      "step": 290
+    },
+    {
+      "epoch": 7.8431372549019605,
+      "grad_norm": 0.8173583149909973,
+      "learning_rate": 0.00019987699691483048,
+      "loss": 0.2595,
+      "step": 300
+    },
+    {
+      "epoch": 8.104575163398692,
+      "grad_norm": 1.2094272375106812,
+      "learning_rate": 0.0001998686618644384,
+      "loss": 0.2696,
+      "step": 310
+    },
+    {
+      "epoch": 8.366013071895425,
+      "grad_norm": 0.8234169483184814,
+      "learning_rate": 0.00019986005377693825,
+      "loss": 0.2446,
+      "step": 320
+    },
+    {
+      "epoch": 8.627450980392156,
+      "grad_norm": 1.0735175609588623,
+      "learning_rate": 0.00019985117267586424,
+      "loss": 0.2667,
+      "step": 330
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 0.8714181184768677,
+      "learning_rate": 0.00019984201858549693,
+      "loss": 0.2737,
+      "step": 340
+    },
+    {
+      "epoch": 9.15032679738562,
+      "grad_norm": 1.0795856714248657,
+      "learning_rate": 0.00019983259153086327,
+      "loss": 0.2782,
+      "step": 350
+    },
+    {
+      "epoch": 9.411764705882353,
+      "grad_norm": 1.0457770824432373,
+      "learning_rate": 0.00019982289153773646,
+      "loss": 0.2415,
+      "step": 360
+    },
+    {
+      "epoch": 9.673202614379084,
+      "grad_norm": 0.8930032849311829,
+      "learning_rate": 0.00019981291863263592,
+      "loss": 0.2492,
+      "step": 370
+    },
+    {
+      "epoch": 9.934640522875817,
+      "grad_norm": 1.0066367387771606,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 0.2467,
+      "step": 380
+    },
+    {
+      "epoch": 10.196078431372548,
+      "grad_norm": 1.148116111755371,
+      "learning_rate": 0.00019979215419632182,
+      "loss": 0.2323,
+      "step": 390
+    },
+    {
+      "epoch": 10.457516339869281,
+      "grad_norm": 1.03926420211792,
+      "learning_rate": 0.00019978136272187747,
+      "loss": 0.237,
+      "step": 400
+    },
+    {
+      "epoch": 10.718954248366012,
+      "grad_norm": 0.9982312917709351,
+      "learning_rate": 0.00019977029844899758,
+      "loss": 0.2597,
+      "step": 410
+    },
+    {
+      "epoch": 10.980392156862745,
+      "grad_norm": 1.2115322351455688,
+      "learning_rate": 0.00019975896140793142,
+      "loss": 0.2409,
+      "step": 420
+    },
+    {
+      "epoch": 11.241830065359476,
+      "grad_norm": 1.4162884950637817,
+      "learning_rate": 0.0001997473516296741,
+      "loss": 0.2381,
+      "step": 430
+    },
+    {
+      "epoch": 11.50326797385621,
+      "grad_norm": 1.3810129165649414,
+      "learning_rate": 0.00019973546914596623,
+      "loss": 0.2491,
+      "step": 440
+    },
+    {
+      "epoch": 11.764705882352942,
+      "grad_norm": 1.0123716592788696,
+      "learning_rate": 0.0001997233139892941,
+      "loss": 0.2401,
+      "step": 450
+    },
+    {
+      "epoch": 12.026143790849673,
+      "grad_norm": 0.9082147479057312,
+      "learning_rate": 0.0001997108861928895,
+      "loss": 0.2382,
+      "step": 460
+    },
+    {
+      "epoch": 12.287581699346406,
+      "grad_norm": 0.9988301992416382,
+      "learning_rate": 0.00019969818579072945,
+      "loss": 0.2217,
+      "step": 470
+    },
+    {
+      "epoch": 12.549019607843137,
+      "grad_norm": 1.2657634019851685,
+      "learning_rate": 0.00019968521281753642,
+      "loss": 0.2417,
+      "step": 480
+    },
+    {
+      "epoch": 12.81045751633987,
+      "grad_norm": 0.8032001852989197,
+      "learning_rate": 0.00019967196730877803,
+      "loss": 0.2309,
+      "step": 490
+    },
+    {
+      "epoch": 13.071895424836601,
+      "grad_norm": 1.2560818195343018,
+      "learning_rate": 0.000199658449300667,
+      "loss": 0.2306,
+      "step": 500
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 0.9680474996566772,
+      "learning_rate": 0.000199644658830161,
+      "loss": 0.217,
+      "step": 510
+    },
+    {
+      "epoch": 13.594771241830065,
+      "grad_norm": 0.8553510904312134,
+      "learning_rate": 0.00019963059593496268,
+      "loss": 0.2362,
+      "step": 520
+    },
+    {
+      "epoch": 13.856209150326798,
+      "grad_norm": 1.1278111934661865,
+      "learning_rate": 0.00019961626065351947,
+      "loss": 0.2252,
+      "step": 530
+    },
+    {
+      "epoch": 14.117647058823529,
+      "grad_norm": 0.8731440305709839,
+      "learning_rate": 0.0001996016530250235,
+      "loss": 0.2101,
+      "step": 540
+    },
+    {
+      "epoch": 14.379084967320262,
+      "grad_norm": 1.0482535362243652,
+      "learning_rate": 0.00019958677308941139,
+      "loss": 0.2113,
+      "step": 550
+    },
+    {
+      "epoch": 14.640522875816993,
+      "grad_norm": 0.8031530380249023,
+      "learning_rate": 0.0001995716208873644,
+      "loss": 0.2285,
+      "step": 560
+    },
+    {
+      "epoch": 14.901960784313726,
+      "grad_norm": 1.2343863248825073,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 0.2225,
+      "step": 570
+    },
+    {
+      "epoch": 15.163398692810457,
+      "grad_norm": 1.141625165939331,
+      "learning_rate": 0.00019954049985041204,
+      "loss": 0.222,
+      "step": 580
+    },
+    {
+      "epoch": 15.42483660130719,
+      "grad_norm": 1.163808822631836,
+      "learning_rate": 0.00019952453110059045,
+      "loss": 0.2099,
+      "step": 590
+    },
+    {
+      "epoch": 15.686274509803921,
+      "grad_norm": 1.0026309490203857,
+      "learning_rate": 0.00019950829025450114,
+      "loss": 0.222,
+      "step": 600
+    },
+    {
+      "epoch": 15.947712418300654,
+      "grad_norm": 0.9590998888015747,
+      "learning_rate": 0.00019949177735654601,
+      "loss": 0.2148,
+      "step": 610
+    },
+    {
+      "epoch": 16.209150326797385,
+      "grad_norm": 1.1316090822219849,
+      "learning_rate": 0.00019947499245187068,
+      "loss": 0.2063,
+      "step": 620
+    },
+    {
+      "epoch": 16.470588235294116,
+      "grad_norm": 1.1160564422607422,
+      "learning_rate": 0.00019945793558636437,
+      "loss": 0.2083,
+      "step": 630
+    },
+    {
+      "epoch": 16.73202614379085,
+      "grad_norm": 1.0182403326034546,
+      "learning_rate": 0.00019944060680666002,
+      "loss": 0.2133,
+      "step": 640
+    },
+    {
+      "epoch": 16.99346405228758,
+      "grad_norm": 0.9575550556182861,
+      "learning_rate": 0.0001994230061601338,
+      "loss": 0.2088,
+      "step": 650
+    },
+    {
+      "epoch": 17.254901960784313,
+      "grad_norm": 1.2370105981826782,
+      "learning_rate": 0.00019940513369490516,
+      "loss": 0.205,
+      "step": 660
+    },
+    {
+      "epoch": 17.516339869281047,
+      "grad_norm": 0.7969275712966919,
+      "learning_rate": 0.00019938698945983676,
+      "loss": 0.2053,
+      "step": 670
+    },
+    {
+      "epoch": 17.77777777777778,
+      "grad_norm": 1.0018616914749146,
+      "learning_rate": 0.0001993685735045343,
+      "loss": 0.2086,
+      "step": 680
+    },
+    {
+      "epoch": 18.03921568627451,
+      "grad_norm": 0.9241987466812134,
+      "learning_rate": 0.00019934988587934623,
+      "loss": 0.2025,
+      "step": 690
+    },
+    {
+      "epoch": 18.30065359477124,
+      "grad_norm": 1.1506460905075073,
+      "learning_rate": 0.00019933092663536382,
+      "loss": 0.1951,
+      "step": 700
+    },
+    {
+      "epoch": 18.562091503267975,
+      "grad_norm": 1.3489614725112915,
+      "learning_rate": 0.00019931169582442095,
+      "loss": 0.2139,
+      "step": 710
+    },
+    {
+      "epoch": 18.823529411764707,
+      "grad_norm": 1.0618153810501099,
+      "learning_rate": 0.00019929219349909392,
+      "loss": 0.2014,
+      "step": 720
+    },
+    {
+      "epoch": 19.084967320261438,
+      "grad_norm": 1.1984943151474,
+      "learning_rate": 0.0001992724197127013,
+      "loss": 0.2084,
+      "step": 730
+    },
+    {
+      "epoch": 19.34640522875817,
+      "grad_norm": 0.9617899656295776,
+      "learning_rate": 0.0001992523745193039,
+      "loss": 0.1913,
+      "step": 740
+    },
+    {
+      "epoch": 19.607843137254903,
+      "grad_norm": 0.951430082321167,
+      "learning_rate": 0.0001992320579737045,
+      "loss": 0.1922,
+      "step": 750
+    },
+    {
+      "epoch": 19.869281045751634,
+      "grad_norm": 1.0308970212936401,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.2028,
+      "step": 760
+    },
+    {
+      "epoch": 20.130718954248366,
+      "grad_norm": 1.298295497894287,
+      "learning_rate": 0.0001991906110488201,
+      "loss": 0.1985,
+      "step": 770
+    },
+    {
+      "epoch": 20.392156862745097,
+      "grad_norm": 0.980803370475769,
+      "learning_rate": 0.0001991694807828494,
+      "loss": 0.2017,
+      "step": 780
+    },
+    {
+      "epoch": 20.65359477124183,
+      "grad_norm": 0.9850193858146667,
+      "learning_rate": 0.000199148079391305,
+      "loss": 0.1808,
+      "step": 790
+    },
+    {
+      "epoch": 20.915032679738562,
+      "grad_norm": 0.8435742855072021,
+      "learning_rate": 0.00019912640693269752,
+      "loss": 0.1991,
+      "step": 800
+    },
+    {
+      "epoch": 21.176470588235293,
+      "grad_norm": 1.2562199831008911,
+      "learning_rate": 0.00019910446346627862,
+      "loss": 0.1902,
+      "step": 810
+    },
+    {
+      "epoch": 21.437908496732025,
+      "grad_norm": 1.2811332941055298,
+      "learning_rate": 0.0001990822490520409,
+      "loss": 0.1895,
+      "step": 820
+    },
+    {
+      "epoch": 21.69934640522876,
+      "grad_norm": 1.185219407081604,
+      "learning_rate": 0.00019905976375071772,
+      "loss": 0.1873,
+      "step": 830
+    },
+    {
+      "epoch": 21.96078431372549,
+      "grad_norm": 1.0126444101333618,
+      "learning_rate": 0.000199037007623783,
+      "loss": 0.2022,
+      "step": 840
+    },
+    {
+      "epoch": 22.22222222222222,
+      "grad_norm": 1.0308198928833008,
+      "learning_rate": 0.00019901398073345118,
+      "loss": 0.186,
+      "step": 850
+    },
+    {
+      "epoch": 22.483660130718953,
+      "grad_norm": 0.9931133389472961,
+      "learning_rate": 0.00019899068314267688,
+      "loss": 0.1892,
+      "step": 860
+    },
+    {
+      "epoch": 22.745098039215687,
+      "grad_norm": 1.1834982633590698,
+      "learning_rate": 0.00019896711491515482,
+      "loss": 0.1866,
+      "step": 870
+    },
+    {
+      "epoch": 23.00653594771242,
+      "grad_norm": 1.0095534324645996,
+      "learning_rate": 0.0001989432761153196,
+      "loss": 0.1877,
+      "step": 880
+    },
+    {
+      "epoch": 23.26797385620915,
+      "grad_norm": 1.0486822128295898,
+      "learning_rate": 0.00019891916680834566,
+      "loss": 0.1848,
+      "step": 890
+    },
+    {
+      "epoch": 23.529411764705884,
+      "grad_norm": 1.0561635494232178,
+      "learning_rate": 0.00019889478706014687,
+      "loss": 0.1825,
+      "step": 900
+    },
+    {
+      "epoch": 23.790849673202615,
+      "grad_norm": 1.0218344926834106,
+      "learning_rate": 0.00019887013693737653,
+      "loss": 0.1781,
+      "step": 910
+    },
+    {
+      "epoch": 24.052287581699346,
+      "grad_norm": 0.907986044883728,
+      "learning_rate": 0.00019884521650742715,
+      "loss": 0.1925,
+      "step": 920
+    },
+    {
+      "epoch": 24.313725490196077,
+      "grad_norm": 1.3268672227859497,
+      "learning_rate": 0.00019882002583843024,
+      "loss": 0.1783,
+      "step": 930
+    },
+    {
+      "epoch": 24.575163398692812,
+      "grad_norm": 0.8679687976837158,
+      "learning_rate": 0.00019879456499925614,
+      "loss": 0.1814,
+      "step": 940
+    },
+    {
+      "epoch": 24.836601307189543,
+      "grad_norm": 1.194421648979187,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.1818,
+      "step": 950
+    },
+    {
+      "epoch": 25.098039215686274,
+      "grad_norm": 1.2440810203552246,
+      "learning_rate": 0.00019874283308955057,
+      "loss": 0.1841,
+      "step": 960
+    },
+    {
+      "epoch": 25.359477124183005,
+      "grad_norm": 0.8485437035560608,
+      "learning_rate": 0.0001987165621604522,
+      "loss": 0.1721,
+      "step": 970
+    },
+    {
+      "epoch": 25.62091503267974,
+      "grad_norm": 1.1203727722167969,
+      "learning_rate": 0.00019869002134404235,
+      "loss": 0.1772,
+      "step": 980
+    },
+    {
+      "epoch": 25.88235294117647,
+      "grad_norm": 1.2701900005340576,
+      "learning_rate": 0.0001986632107128826,
+      "loss": 0.1823,
+      "step": 990
+    },
+    {
+      "epoch": 26.143790849673202,
+      "grad_norm": 0.7330073118209839,
+      "learning_rate": 0.00019863613034027224,
+      "loss": 0.1762,
+      "step": 1000
+    },
+    {
+      "epoch": 26.405228758169933,
+      "grad_norm": 1.0160173177719116,
+      "learning_rate": 0.0001986087803002479,
+      "loss": 0.1748,
+      "step": 1010
+    },
+    {
+      "epoch": 26.666666666666668,
+      "grad_norm": 1.2125264406204224,
+      "learning_rate": 0.00019858116066758362,
+      "loss": 0.1818,
+      "step": 1020
+    },
+    {
+      "epoch": 26.9281045751634,
+      "grad_norm": 1.3150966167449951,
+      "learning_rate": 0.00019855327151779042,
+      "loss": 0.1791,
+      "step": 1030
+    },
+    {
+      "epoch": 27.18954248366013,
+      "grad_norm": 0.9498183131217957,
+      "learning_rate": 0.00019852511292711608,
+      "loss": 0.1746,
+      "step": 1040
+    },
+    {
+      "epoch": 27.45098039215686,
+      "grad_norm": 1.101545810699463,
+      "learning_rate": 0.0001984966849725452,
+      "loss": 0.17,
+      "step": 1050
+    },
+    {
+      "epoch": 27.712418300653596,
+      "grad_norm": 0.9787285923957825,
+      "learning_rate": 0.00019846798773179866,
+      "loss": 0.1759,
+      "step": 1060
+    },
+    {
+      "epoch": 27.973856209150327,
+      "grad_norm": 0.9753018021583557,
+      "learning_rate": 0.00019843902128333367,
+      "loss": 0.1823,
+      "step": 1070
+    },
+    {
+      "epoch": 28.235294117647058,
+      "grad_norm": 1.1331627368927002,
+      "learning_rate": 0.0001984097857063434,
+      "loss": 0.1682,
+      "step": 1080
+    },
+    {
+      "epoch": 28.49673202614379,
+      "grad_norm": 1.199524998664856,
+      "learning_rate": 0.00019838028108075671,
+      "loss": 0.1735,
+      "step": 1090
+    },
+    {
+      "epoch": 28.758169934640524,
+      "grad_norm": 0.83395916223526,
+      "learning_rate": 0.00019835050748723824,
+      "loss": 0.1683,
+      "step": 1100
+    },
+    {
+      "epoch": 29.019607843137255,
+      "grad_norm": 0.954365611076355,
+      "learning_rate": 0.0001983204650071878,
+      "loss": 0.1764,
+      "step": 1110
+    },
+    {
+      "epoch": 29.281045751633986,
+      "grad_norm": 1.1084930896759033,
+      "learning_rate": 0.00019829015372274038,
+      "loss": 0.1663,
+      "step": 1120
+    },
+    {
+      "epoch": 29.54248366013072,
+      "grad_norm": 0.9061603546142578,
+      "learning_rate": 0.00019825957371676588,
+      "loss": 0.1618,
+      "step": 1130
+    },
+    {
+      "epoch": 29.80392156862745,
+      "grad_norm": 1.2576953172683716,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 0.1773,
+      "step": 1140
+    },
+    {
+      "epoch": 30.065359477124183,
+      "grad_norm": 0.680230438709259,
+      "learning_rate": 0.0001981976078753884,
+      "loss": 0.1749,
+      "step": 1150
+    },
+    {
+      "epoch": 30.326797385620914,
+      "grad_norm": 1.0407228469848633,
+      "learning_rate": 0.0001981662222093976,
+      "loss": 0.1608,
+      "step": 1160
+    },
+    {
+      "epoch": 30.58823529411765,
+      "grad_norm": 1.0547195672988892,
+      "learning_rate": 0.0001981345681607038,
+      "loss": 0.1626,
+      "step": 1170
+    },
+    {
+      "epoch": 30.84967320261438,
+      "grad_norm": 1.2236053943634033,
+      "learning_rate": 0.00019810264581584787,
+      "loss": 0.1701,
+      "step": 1180
+    },
+    {
+      "epoch": 31.11111111111111,
+      "grad_norm": 1.0089291334152222,
+      "learning_rate": 0.00019807045526210436,
+      "loss": 0.1679,
+      "step": 1190
+    },
+    {
+      "epoch": 31.372549019607842,
+      "grad_norm": 1.067421555519104,
+      "learning_rate": 0.00019803799658748094,
+      "loss": 0.1614,
+      "step": 1200
+    },
+    {
+      "epoch": 31.633986928104576,
+      "grad_norm": 1.1328610181808472,
+      "learning_rate": 0.00019800526988071843,
+      "loss": 0.1691,
+      "step": 1210
+    },
+    {
+      "epoch": 31.895424836601308,
+      "grad_norm": 1.216069221496582,
+      "learning_rate": 0.0001979722752312904,
+      "loss": 0.1674,
+      "step": 1220
+    },
+    {
+      "epoch": 32.15686274509804,
+      "grad_norm": 0.9567670822143555,
+      "learning_rate": 0.00019793901272940293,
+      "loss": 0.1592,
+      "step": 1230
+    },
+    {
+      "epoch": 32.41830065359477,
+      "grad_norm": 0.9017887115478516,
+      "learning_rate": 0.00019790548246599447,
+      "loss": 0.1598,
+      "step": 1240
+    },
+    {
+      "epoch": 32.6797385620915,
+      "grad_norm": 1.1113390922546387,
+      "learning_rate": 0.00019787168453273544,
+      "loss": 0.1659,
+      "step": 1250
+    },
+    {
+      "epoch": 32.94117647058823,
+      "grad_norm": 1.2047311067581177,
+      "learning_rate": 0.00019783761902202813,
+      "loss": 0.1682,
+      "step": 1260
+    },
+    {
+      "epoch": 33.20261437908497,
+      "grad_norm": 1.0487027168273926,
+      "learning_rate": 0.00019780328602700636,
+      "loss": 0.16,
+      "step": 1270
+    },
+    {
+      "epoch": 33.4640522875817,
+      "grad_norm": 1.0085713863372803,
+      "learning_rate": 0.00019776868564153516,
+      "loss": 0.1544,
+      "step": 1280
+    },
+    {
+      "epoch": 33.72549019607843,
+      "grad_norm": 1.049383521080017,
+      "learning_rate": 0.0001977338179602107,
+      "loss": 0.1687,
+      "step": 1290
+    },
+    {
+      "epoch": 33.98692810457516,
+      "grad_norm": 1.0386006832122803,
+      "learning_rate": 0.00019769868307835994,
+      "loss": 0.1637,
+      "step": 1300
+    },
+    {
+      "epoch": 34.248366013071895,
+      "grad_norm": 1.0458704233169556,
+      "learning_rate": 0.00019766328109204024,
+      "loss": 0.1578,
+      "step": 1310
+    },
+    {
+      "epoch": 34.509803921568626,
+      "grad_norm": 1.0181804895401,
+      "learning_rate": 0.00019762761209803927,
+      "loss": 0.1556,
+      "step": 1320
+    },
+    {
+      "epoch": 34.77124183006536,
+      "grad_norm": 0.8453966975212097,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 0.1594,
+      "step": 1330
+    },
+    {
+      "epoch": 35.032679738562095,
+      "grad_norm": 0.8666368126869202,
+      "learning_rate": 0.00019755547347779403,
+      "loss": 0.1639,
+      "step": 1340
+    },
+    {
+      "epoch": 35.294117647058826,
+      "grad_norm": 0.8416698575019836,
+      "learning_rate": 0.000197519004048774,
+      "loss": 0.1495,
+      "step": 1350
+    },
+    {
+      "epoch": 35.55555555555556,
+      "grad_norm": 1.3281161785125732,
+      "learning_rate": 0.0001974822680065206,
+      "loss": 0.1574,
+      "step": 1360
+    },
+    {
+      "epoch": 35.81699346405229,
+      "grad_norm": 0.953856348991394,
+      "learning_rate": 0.00019744526545146886,
+      "loss": 0.1614,
+      "step": 1370
+    },
+    {
+      "epoch": 36.07843137254902,
+      "grad_norm": 0.6768958568572998,
+      "learning_rate": 0.00019740799648478233,
+      "loss": 0.1571,
+      "step": 1380
+    },
+    {
+      "epoch": 36.33986928104575,
+      "grad_norm": 1.09906005859375,
+      "learning_rate": 0.0001973704612083529,
+      "loss": 0.1551,
+      "step": 1390
+    },
+    {
+      "epoch": 36.60130718954248,
+      "grad_norm": 1.172494649887085,
+      "learning_rate": 0.0001973326597248006,
+      "loss": 0.1597,
+      "step": 1400
+    },
+    {
+      "epoch": 36.86274509803921,
+      "grad_norm": 1.0662363767623901,
+      "learning_rate": 0.00019729459213747326,
+      "loss": 0.1625,
+      "step": 1410
+    },
+    {
+      "epoch": 37.12418300653595,
+      "grad_norm": 0.7910565137863159,
+      "learning_rate": 0.00019725625855044617,
+      "loss": 0.1504,
+      "step": 1420
+    },
+    {
+      "epoch": 37.38562091503268,
+      "grad_norm": 1.1636861562728882,
+      "learning_rate": 0.00019721765906852197,
+      "loss": 0.1508,
+      "step": 1430
+    },
+    {
+      "epoch": 37.64705882352941,
+      "grad_norm": 1.0530037879943848,
+      "learning_rate": 0.00019717879379723012,
+      "loss": 0.1543,
+      "step": 1440
+    },
+    {
+      "epoch": 37.908496732026144,
+      "grad_norm": 1.4254627227783203,
+      "learning_rate": 0.00019713966284282678,
+      "loss": 0.1624,
+      "step": 1450
+    },
+    {
+      "epoch": 38.169934640522875,
+      "grad_norm": 1.1266350746154785,
+      "learning_rate": 0.0001971002663122945,
+      "loss": 0.1505,
+      "step": 1460
+    },
+    {
+      "epoch": 38.431372549019606,
+      "grad_norm": 0.8052533864974976,
+      "learning_rate": 0.00019706060431334187,
+      "loss": 0.1473,
+      "step": 1470
+    },
+    {
+      "epoch": 38.69281045751634,
+      "grad_norm": 1.1820138692855835,
+      "learning_rate": 0.00019702067695440332,
+      "loss": 0.1512,
+      "step": 1480
+    },
+    {
+      "epoch": 38.95424836601307,
+      "grad_norm": 1.0051476955413818,
+      "learning_rate": 0.0001969804843446387,
+      "loss": 0.164,
+      "step": 1490
+    },
+    {
+      "epoch": 39.21568627450981,
+      "grad_norm": 1.0447814464569092,
+      "learning_rate": 0.00019694002659393305,
+      "loss": 0.1503,
+      "step": 1500
+    },
+    {
+      "epoch": 39.47712418300654,
+      "grad_norm": 0.8165223598480225,
+      "learning_rate": 0.00019689930381289634,
+      "loss": 0.1525,
+      "step": 1510
+    },
+    {
+      "epoch": 39.73856209150327,
+      "grad_norm": 0.9069657325744629,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.1539,
+      "step": 1520
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 1.4593009948730469,
+      "learning_rate": 0.00019681706360589216,
+      "loss": 0.1572,
+      "step": 1530
+    },
+    {
+      "epoch": 40.26143790849673,
+      "grad_norm": 1.3013628721237183,
+      "learning_rate": 0.00019677554640476624,
+      "loss": 0.1459,
+      "step": 1540
+    },
+    {
+      "epoch": 40.52287581699346,
+      "grad_norm": 1.1023391485214233,
+      "learning_rate": 0.00019673376462299184,
+      "loss": 0.1526,
+      "step": 1550
+    },
+    {
+      "epoch": 40.78431372549019,
+      "grad_norm": 1.0940853357315063,
+      "learning_rate": 0.00019669171837479873,
+      "loss": 0.1581,
+      "step": 1560
+    },
+    {
+      "epoch": 41.04575163398693,
+      "grad_norm": 0.940166175365448,
+      "learning_rate": 0.00019664940777513974,
+      "loss": 0.1492,
+      "step": 1570
+    },
+    {
+      "epoch": 41.30718954248366,
+      "grad_norm": 0.9569455981254578,
+      "learning_rate": 0.00019660683293969041,
+      "loss": 0.1453,
+      "step": 1580
+    },
+    {
+      "epoch": 41.568627450980394,
+      "grad_norm": 0.7959948182106018,
+      "learning_rate": 0.0001965639939848488,
+      "loss": 0.1505,
+      "step": 1590
+    },
+    {
+      "epoch": 41.830065359477125,
+      "grad_norm": 0.991958498954773,
+      "learning_rate": 0.00019652089102773488,
+      "loss": 0.1518,
+      "step": 1600
+    },
+    {
+      "epoch": 42.091503267973856,
+      "grad_norm": 1.0243059396743774,
+      "learning_rate": 0.00019647752418619054,
+      "loss": 0.1515,
+      "step": 1610
+    },
+    {
+      "epoch": 42.35294117647059,
+      "grad_norm": 0.826440155506134,
+      "learning_rate": 0.00019643389357877907,
+      "loss": 0.1465,
+      "step": 1620
+    },
+    {
+      "epoch": 42.61437908496732,
+      "grad_norm": 0.8773382902145386,
+      "learning_rate": 0.00019638999932478486,
+      "loss": 0.1468,
+      "step": 1630
+    },
+    {
+      "epoch": 42.87581699346405,
+      "grad_norm": 0.7697790861129761,
+      "learning_rate": 0.00019634584154421317,
+      "loss": 0.1442,
+      "step": 1640
+    },
+    {
+      "epoch": 43.13725490196079,
+      "grad_norm": 0.7857301831245422,
+      "learning_rate": 0.00019630142035778964,
+      "loss": 0.1473,
+      "step": 1650
+    },
+    {
+      "epoch": 43.39869281045752,
+      "grad_norm": 0.8533034920692444,
+      "learning_rate": 0.00019625673588696008,
+      "loss": 0.1453,
+      "step": 1660
+    },
+    {
+      "epoch": 43.66013071895425,
+      "grad_norm": 0.8283002376556396,
+      "learning_rate": 0.0001962117882538902,
+      "loss": 0.1469,
+      "step": 1670
+    },
+    {
+      "epoch": 43.92156862745098,
+      "grad_norm": 1.0466629266738892,
+      "learning_rate": 0.00019616657758146503,
+      "loss": 0.1501,
+      "step": 1680
+    },
+    {
+      "epoch": 44.18300653594771,
+      "grad_norm": 0.7576202750205994,
+      "learning_rate": 0.00019612110399328892,
+      "loss": 0.1438,
+      "step": 1690
+    },
+    {
+      "epoch": 44.44444444444444,
+      "grad_norm": 0.8621203303337097,
+      "learning_rate": 0.00019607536761368484,
+      "loss": 0.1382,
+      "step": 1700
+    },
+    {
+      "epoch": 44.705882352941174,
+      "grad_norm": 0.864361047744751,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 0.1488,
+      "step": 1710
+    },
+    {
+      "epoch": 44.967320261437905,
+      "grad_norm": 1.3524795770645142,
+      "learning_rate": 0.00019598310698107702,
+      "loss": 0.1501,
+      "step": 1720
+    },
+    {
+      "epoch": 45.22875816993464,
+      "grad_norm": 0.8673699498176575,
+      "learning_rate": 0.00019593658298031034,
+      "loss": 0.1446,
+      "step": 1730
+    },
+    {
+      "epoch": 45.490196078431374,
+      "grad_norm": 0.7573821544647217,
+      "learning_rate": 0.0001958897966925891,
+      "loss": 0.1404,
+      "step": 1740
+    },
+    {
+      "epoch": 45.751633986928105,
+      "grad_norm": 1.00885009765625,
+      "learning_rate": 0.0001958427482458253,
+      "loss": 0.1499,
+      "step": 1750
+    },
+    {
+      "epoch": 46.01307189542484,
+      "grad_norm": 1.3049975633621216,
+      "learning_rate": 0.0001957954377686475,
+      "loss": 0.1494,
+      "step": 1760
+    },
+    {
+      "epoch": 46.27450980392157,
+      "grad_norm": 0.8838759660720825,
+      "learning_rate": 0.00019574786539040077,
+      "loss": 0.1401,
+      "step": 1770
+    },
+    {
+      "epoch": 46.5359477124183,
+      "grad_norm": 0.7831525206565857,
+      "learning_rate": 0.00019570003124114619,
+      "loss": 0.1471,
+      "step": 1780
+    },
+    {
+      "epoch": 46.79738562091503,
+      "grad_norm": 0.7171520590782166,
+      "learning_rate": 0.00019565193545166052,
+      "loss": 0.1414,
+      "step": 1790
+    },
+    {
+      "epoch": 47.05882352941177,
+      "grad_norm": 0.7733674645423889,
+      "learning_rate": 0.00019560357815343577,
+      "loss": 0.1465,
+      "step": 1800
+    },
+    {
+      "epoch": 47.3202614379085,
+      "grad_norm": 0.833474338054657,
+      "learning_rate": 0.00019555495947867895,
+      "loss": 0.1402,
+      "step": 1810
+    },
+    {
+      "epoch": 47.58169934640523,
+      "grad_norm": 0.8420645594596863,
+      "learning_rate": 0.0001955060795603117,
+      "loss": 0.1412,
+      "step": 1820
+    },
+    {
+      "epoch": 47.84313725490196,
+      "grad_norm": 0.9879257678985596,
+      "learning_rate": 0.00019545693853196983,
+      "loss": 0.1444,
+      "step": 1830
+    },
+    {
+      "epoch": 48.10457516339869,
+      "grad_norm": 0.9052282571792603,
+      "learning_rate": 0.000195407536528003,
+      "loss": 0.1433,
+      "step": 1840
+    },
+    {
+      "epoch": 48.36601307189542,
+      "grad_norm": 1.194151759147644,
+      "learning_rate": 0.00019535787368347442,
+      "loss": 0.1419,
+      "step": 1850
+    },
+    {
+      "epoch": 48.627450980392155,
+      "grad_norm": 1.0415573120117188,
+      "learning_rate": 0.00019530795013416046,
+      "loss": 0.1413,
+      "step": 1860
+    },
+    {
+      "epoch": 48.888888888888886,
+      "grad_norm": 0.9441238641738892,
+      "learning_rate": 0.00019525776601655014,
+      "loss": 0.1473,
+      "step": 1870
+    },
+    {
+      "epoch": 49.150326797385624,
+      "grad_norm": 0.7534403800964355,
+      "learning_rate": 0.00019520732146784491,
+      "loss": 0.1386,
+      "step": 1880
+    },
+    {
+      "epoch": 49.411764705882355,
+      "grad_norm": 1.0693260431289673,
+      "learning_rate": 0.0001951566166259583,
+      "loss": 0.1419,
+      "step": 1890
+    },
+    {
+      "epoch": 49.673202614379086,
+      "grad_norm": 0.9667706489562988,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.1411,
+      "step": 1900
+    },
+    {
+      "epoch": 49.93464052287582,
+      "grad_norm": 0.8170807361602783,
+      "learning_rate": 0.0001950544266178525,
+      "loss": 0.146,
+      "step": 1910
+    },
+    {
+      "epoch": 50.19607843137255,
+      "grad_norm": 0.820950448513031,
+      "learning_rate": 0.00019500294173101687,
+      "loss": 0.1409,
+      "step": 1920
+    },
+    {
+      "epoch": 50.45751633986928,
+      "grad_norm": 0.9750927686691284,
+      "learning_rate": 0.00019495119710976626,
+      "loss": 0.1381,
+      "step": 1930
+    },
+    {
+      "epoch": 50.71895424836601,
+      "grad_norm": 0.9925455451011658,
+      "learning_rate": 0.00019489919289556845,
+      "loss": 0.1455,
+      "step": 1940
+    },
+    {
+      "epoch": 50.98039215686274,
+      "grad_norm": 0.7548269033432007,
+      "learning_rate": 0.00019484692923060095,
+      "loss": 0.1399,
+      "step": 1950
+    },
+    {
+      "epoch": 51.24183006535948,
+      "grad_norm": 0.7531190514564514,
+      "learning_rate": 0.0001947944062577507,
+      "loss": 0.1343,
+      "step": 1960
+    },
+    {
+      "epoch": 51.50326797385621,
+      "grad_norm": 0.9674361348152161,
+      "learning_rate": 0.0001947416241206134,
+      "loss": 0.1382,
+      "step": 1970
+    },
+    {
+      "epoch": 51.76470588235294,
+      "grad_norm": 0.9091185331344604,
+      "learning_rate": 0.0001946885829634935,
+      "loss": 0.1445,
+      "step": 1980
+    },
+    {
+      "epoch": 52.02614379084967,
+      "grad_norm": 0.987326979637146,
+      "learning_rate": 0.00019463528293140345,
+      "loss": 0.1453,
+      "step": 1990
+    },
+    {
+      "epoch": 52.287581699346404,
+      "grad_norm": 0.6855144500732422,
+      "learning_rate": 0.00019458172417006347,
+      "loss": 0.1349,
+      "step": 2000
+    },
+    {
+      "epoch": 52.549019607843135,
+      "grad_norm": 0.7305368781089783,
+      "learning_rate": 0.00019452790682590124,
+      "loss": 0.1397,
+      "step": 2010
+    },
+    {
+      "epoch": 52.810457516339866,
+      "grad_norm": 0.9413096904754639,
+      "learning_rate": 0.00019447383104605125,
+      "loss": 0.1446,
+      "step": 2020
+    },
+    {
+      "epoch": 53.071895424836605,
+      "grad_norm": 0.7378990650177002,
+      "learning_rate": 0.0001944194969783547,
+      "loss": 0.1409,
+      "step": 2030
+    },
+    {
+      "epoch": 53.333333333333336,
+      "grad_norm": 0.6865296959877014,
+      "learning_rate": 0.00019436490477135878,
+      "loss": 0.1352,
+      "step": 2040
+    },
+    {
+      "epoch": 53.59477124183007,
+      "grad_norm": 0.71687912940979,
+      "learning_rate": 0.00019431005457431653,
+      "loss": 0.1423,
+      "step": 2050
+    },
+    {
+      "epoch": 53.8562091503268,
+      "grad_norm": 0.7911390662193298,
+      "learning_rate": 0.0001942549465371863,
+      "loss": 0.1374,
+      "step": 2060
+    },
+    {
+      "epoch": 54.11764705882353,
+      "grad_norm": 0.8051833510398865,
+      "learning_rate": 0.00019419958081063138,
+      "loss": 0.1363,
+      "step": 2070
+    },
+    {
+      "epoch": 54.37908496732026,
+      "grad_norm": 0.6623182892799377,
+      "learning_rate": 0.00019414395754601947,
+      "loss": 0.1375,
+      "step": 2080
+    },
+    {
+      "epoch": 54.64052287581699,
+      "grad_norm": 0.8147381544113159,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 0.1331,
+      "step": 2090
+    },
+    {
+      "epoch": 54.90196078431372,
+      "grad_norm": 0.7623370885848999,
+      "learning_rate": 0.00019403193901161613,
+      "loss": 0.14,
+      "step": 2100
+    },
+    {
+      "epoch": 55.16339869281046,
+      "grad_norm": 0.8199501037597656,
+      "learning_rate": 0.00019397554404807906,
+      "loss": 0.1389,
+      "step": 2110
+    },
+    {
+      "epoch": 55.42483660130719,
+      "grad_norm": 1.0729644298553467,
+      "learning_rate": 0.00019391889215899299,
+      "loss": 0.1376,
+      "step": 2120
+    },
+    {
+      "epoch": 55.68627450980392,
+      "grad_norm": 0.6667032241821289,
+      "learning_rate": 0.00019386198349924207,
+      "loss": 0.1359,
+      "step": 2130
+    },
+    {
+      "epoch": 55.947712418300654,
+      "grad_norm": 0.9523484706878662,
+      "learning_rate": 0.00019380481822441235,
+      "loss": 0.1423,
+      "step": 2140
+    },
+    {
+      "epoch": 56.209150326797385,
+      "grad_norm": 0.8002906441688538,
+      "learning_rate": 0.00019374739649079153,
+      "loss": 0.1315,
+      "step": 2150
+    },
+    {
+      "epoch": 56.470588235294116,
+      "grad_norm": 0.6775190830230713,
+      "learning_rate": 0.00019368971845536845,
+      "loss": 0.1342,
+      "step": 2160
+    },
+    {
+      "epoch": 56.73202614379085,
+      "grad_norm": 0.870460569858551,
+      "learning_rate": 0.00019363178427583266,
+      "loss": 0.1398,
+      "step": 2170
+    },
+    {
+      "epoch": 56.99346405228758,
+      "grad_norm": 0.8694108128547668,
+      "learning_rate": 0.000193573594110574,
+      "loss": 0.1453,
+      "step": 2180
+    },
+    {
+      "epoch": 57.254901960784316,
+      "grad_norm": 1.0844203233718872,
+      "learning_rate": 0.00019351514811868207,
+      "loss": 0.1336,
+      "step": 2190
+    },
+    {
+      "epoch": 57.51633986928105,
+      "grad_norm": 0.7382115125656128,
+      "learning_rate": 0.0001934564464599461,
+      "loss": 0.1379,
+      "step": 2200
+    },
+    {
+      "epoch": 57.77777777777778,
+      "grad_norm": 0.996875524520874,
+      "learning_rate": 0.0001933974892948541,
+      "loss": 0.1353,
+      "step": 2210
+    },
+    {
+      "epoch": 58.03921568627451,
+      "grad_norm": 0.7248476147651672,
+      "learning_rate": 0.0001933382767845928,
+      "loss": 0.1386,
+      "step": 2220
+    },
+    {
+      "epoch": 58.30065359477124,
+      "grad_norm": 0.671381413936615,
+      "learning_rate": 0.00019327880909104683,
+      "loss": 0.1347,
+      "step": 2230
+    },
+    {
+      "epoch": 58.56209150326797,
+      "grad_norm": 0.5566993355751038,
+      "learning_rate": 0.00019321908637679865,
+      "loss": 0.1365,
+      "step": 2240
+    },
+    {
+      "epoch": 58.8235294117647,
+      "grad_norm": 0.8637556433677673,
+      "learning_rate": 0.0001931591088051279,
+      "loss": 0.1344,
+      "step": 2250
+    },
+    {
+      "epoch": 59.08496732026144,
+      "grad_norm": 0.6594296097755432,
+      "learning_rate": 0.00019309887654001096,
+      "loss": 0.1359,
+      "step": 2260
+    },
+    {
+      "epoch": 59.34640522875817,
+      "grad_norm": 0.8212260603904724,
+      "learning_rate": 0.0001930383897461205,
+      "loss": 0.1286,
+      "step": 2270
+    },
+    {
+      "epoch": 59.6078431372549,
+      "grad_norm": 0.7467018365859985,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.1379,
+      "step": 2280
+    },
+    {
+      "epoch": 59.869281045751634,
+      "grad_norm": 0.7997498512268066,
+      "learning_rate": 0.0001929166532341889,
+      "loss": 0.1379,
+      "step": 2290
+    },
+    {
+      "epoch": 60.130718954248366,
+      "grad_norm": 0.5982789993286133,
+      "learning_rate": 0.00019285540384897073,
+      "loss": 0.1323,
+      "step": 2300
+    },
+    {
+      "epoch": 60.3921568627451,
+      "grad_norm": 0.8748401403427124,
+      "learning_rate": 0.00019279390060062407,
+      "loss": 0.1317,
+      "step": 2310
+    },
+    {
+      "epoch": 60.65359477124183,
+      "grad_norm": 0.965622067451477,
+      "learning_rate": 0.00019273214365729655,
+      "loss": 0.1355,
+      "step": 2320
+    },
+    {
+      "epoch": 60.91503267973856,
+      "grad_norm": 0.5320419073104858,
+      "learning_rate": 0.0001926701331878292,
+      "loss": 0.1383,
+      "step": 2330
+    },
+    {
+      "epoch": 61.1764705882353,
+      "grad_norm": 0.8016908764839172,
+      "learning_rate": 0.00019260786936175635,
+      "loss": 0.1356,
+      "step": 2340
+    },
+    {
+      "epoch": 61.43790849673203,
+      "grad_norm": 0.5881072878837585,
+      "learning_rate": 0.00019254535234930486,
+      "loss": 0.1344,
+      "step": 2350
+    },
+    {
+      "epoch": 61.69934640522876,
+      "grad_norm": 0.7665542960166931,
+      "learning_rate": 0.00019248258232139388,
+      "loss": 0.1352,
+      "step": 2360
+    },
+    {
+      "epoch": 61.96078431372549,
+      "grad_norm": 0.709434449672699,
+      "learning_rate": 0.0001924195594496343,
+      "loss": 0.1352,
+      "step": 2370
+    },
+    {
+      "epoch": 62.22222222222222,
+      "grad_norm": 0.5670029520988464,
+      "learning_rate": 0.00019235628390632822,
+      "loss": 0.1282,
+      "step": 2380
+    },
+    {
+      "epoch": 62.48366013071895,
+      "grad_norm": 0.9107353091239929,
+      "learning_rate": 0.0001922927558644685,
+      "loss": 0.1294,
+      "step": 2390
+    },
+    {
+      "epoch": 62.745098039215684,
+      "grad_norm": 0.6936364769935608,
+      "learning_rate": 0.00019222897549773848,
+      "loss": 0.1342,
+      "step": 2400
+    },
+    {
+      "epoch": 63.00653594771242,
+      "grad_norm": 0.7786808013916016,
+      "learning_rate": 0.0001921649429805112,
+      "loss": 0.1399,
+      "step": 2410
+    },
+    {
+      "epoch": 63.26797385620915,
+      "grad_norm": 0.6474490165710449,
+      "learning_rate": 0.00019210065848784913,
+      "loss": 0.1266,
+      "step": 2420
+    },
+    {
+      "epoch": 63.529411764705884,
+      "grad_norm": 0.8766961097717285,
+      "learning_rate": 0.00019203612219550358,
+      "loss": 0.1319,
+      "step": 2430
+    },
+    {
+      "epoch": 63.790849673202615,
+      "grad_norm": 0.8006402254104614,
+      "learning_rate": 0.00019197133427991436,
+      "loss": 0.1338,
+      "step": 2440
+    },
+    {
+      "epoch": 64.05228758169935,
+      "grad_norm": 0.7504803538322449,
+      "learning_rate": 0.00019190629491820912,
+      "loss": 0.1408,
+      "step": 2450
+    },
+    {
+      "epoch": 64.31372549019608,
+      "grad_norm": 0.814525842666626,
+      "learning_rate": 0.000191841004288203,
+      "loss": 0.1326,
+      "step": 2460
+    },
+    {
+      "epoch": 64.57516339869281,
+      "grad_norm": 0.8137060403823853,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 0.1316,
+      "step": 2470
+    },
+    {
+      "epoch": 64.83660130718954,
+      "grad_norm": 0.7670857310295105,
+      "learning_rate": 0.000191709669937983,
+      "loss": 0.1348,
+      "step": 2480
+    },
+    {
+      "epoch": 65.09803921568627,
+      "grad_norm": 0.646980345249176,
+      "learning_rate": 0.00019164362657683222,
+      "loss": 0.1333,
+      "step": 2490
+    },
+    {
+      "epoch": 65.359477124183,
+      "grad_norm": 1.1511456966400146,
+      "learning_rate": 0.00019157733266550575,
+      "loss": 0.1341,
+      "step": 2500
+    },
+    {
+      "epoch": 65.62091503267973,
+      "grad_norm": 0.6792668104171753,
+      "learning_rate": 0.00019151078838524867,
+      "loss": 0.1306,
+      "step": 2510
+    },
+    {
+      "epoch": 65.88235294117646,
+      "grad_norm": 0.8312354683876038,
+      "learning_rate": 0.00019144399391799043,
+      "loss": 0.1354,
+      "step": 2520
+    },
+    {
+      "epoch": 66.14379084967321,
+      "grad_norm": 0.7580569386482239,
+      "learning_rate": 0.00019137694944634464,
+      "loss": 0.1333,
+      "step": 2530
+    },
+    {
+      "epoch": 66.40522875816994,
+      "grad_norm": 0.7622929811477661,
+      "learning_rate": 0.0001913096551536083,
+      "loss": 0.1302,
+      "step": 2540
+    },
+    {
+      "epoch": 66.66666666666667,
+      "grad_norm": 0.9241070747375488,
+      "learning_rate": 0.00019124211122376137,
+      "loss": 0.131,
+      "step": 2550
+    },
+    {
+      "epoch": 66.9281045751634,
+      "grad_norm": 0.94864821434021,
+      "learning_rate": 0.00019117431784146645,
+      "loss": 0.1374,
+      "step": 2560
+    },
+    {
+      "epoch": 67.18954248366013,
+      "grad_norm": 0.6629276871681213,
+      "learning_rate": 0.00019110627519206805,
+      "loss": 0.1281,
+      "step": 2570
+    },
+    {
+      "epoch": 67.45098039215686,
+      "grad_norm": 0.868118941783905,
+      "learning_rate": 0.00019103798346159213,
+      "loss": 0.1308,
+      "step": 2580
+    },
+    {
+      "epoch": 67.7124183006536,
+      "grad_norm": 0.7864583730697632,
+      "learning_rate": 0.00019096944283674571,
+      "loss": 0.1332,
+      "step": 2590
+    },
+    {
+      "epoch": 67.97385620915033,
+      "grad_norm": 0.7753363251686096,
+      "learning_rate": 0.00019090065350491626,
+      "loss": 0.136,
+      "step": 2600
+    },
+    {
+      "epoch": 68.23529411764706,
+      "grad_norm": 0.4261542856693268,
+      "learning_rate": 0.00019083161565417115,
+      "loss": 0.1266,
+      "step": 2610
+    },
+    {
+      "epoch": 68.49673202614379,
+      "grad_norm": 0.7071658968925476,
+      "learning_rate": 0.00019076232947325722,
+      "loss": 0.1277,
+      "step": 2620
+    },
+    {
+      "epoch": 68.75816993464052,
+      "grad_norm": 0.913390040397644,
+      "learning_rate": 0.00019069279515160025,
+      "loss": 0.1335,
+      "step": 2630
+    },
+    {
+      "epoch": 69.01960784313725,
+      "grad_norm": 0.8911900520324707,
+      "learning_rate": 0.00019062301287930446,
+      "loss": 0.1352,
+      "step": 2640
+    },
+    {
+      "epoch": 69.28104575163398,
+      "grad_norm": 0.7245123982429504,
+      "learning_rate": 0.00019055298284715192,
+      "loss": 0.1286,
+      "step": 2650
+    },
+    {
+      "epoch": 69.54248366013071,
+      "grad_norm": 0.7637701034545898,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 0.1325,
+      "step": 2660
+    },
+    {
+      "epoch": 69.80392156862744,
+      "grad_norm": 0.5611773133277893,
+      "learning_rate": 0.00019041218026979095,
+      "loss": 0.132,
+      "step": 2670
+    },
+    {
+      "epoch": 70.06535947712419,
+      "grad_norm": 0.6525811553001404,
+      "learning_rate": 0.0001903414081095315,
+      "loss": 0.1303,
+      "step": 2680
+    },
+    {
+      "epoch": 70.32679738562092,
+      "grad_norm": 0.5672200322151184,
+      "learning_rate": 0.0001902703889593119,
+      "loss": 0.1253,
+      "step": 2690
+    },
+    {
+      "epoch": 70.58823529411765,
+      "grad_norm": 0.698940098285675,
+      "learning_rate": 0.00019019912301329592,
+      "loss": 0.1261,
+      "step": 2700
+    },
+    {
+      "epoch": 70.84967320261438,
+      "grad_norm": 0.616579532623291,
+      "learning_rate": 0.0001901276104663218,
+      "loss": 0.1354,
+      "step": 2710
+    },
+    {
+      "epoch": 71.11111111111111,
+      "grad_norm": 0.5749228000640869,
+      "learning_rate": 0.00019005585151390223,
+      "loss": 0.1297,
+      "step": 2720
+    },
+    {
+      "epoch": 71.37254901960785,
+      "grad_norm": 0.9096235632896423,
+      "learning_rate": 0.00018998384635222334,
+      "loss": 0.1263,
+      "step": 2730
+    },
+    {
+      "epoch": 71.63398692810458,
+      "grad_norm": 0.8338767290115356,
+      "learning_rate": 0.0001899115951781446,
+      "loss": 0.1301,
+      "step": 2740
+    },
+    {
+      "epoch": 71.89542483660131,
+      "grad_norm": 0.7974913716316223,
+      "learning_rate": 0.0001898390981891979,
+      "loss": 0.1337,
+      "step": 2750
+    },
+    {
+      "epoch": 72.15686274509804,
+      "grad_norm": 0.5519678592681885,
+      "learning_rate": 0.00018976635558358722,
+      "loss": 0.1287,
+      "step": 2760
+    },
+    {
+      "epoch": 72.41830065359477,
+      "grad_norm": 0.7348606586456299,
+      "learning_rate": 0.00018969336756018815,
+      "loss": 0.1272,
+      "step": 2770
+    },
+    {
+      "epoch": 72.6797385620915,
+      "grad_norm": 0.6163354516029358,
+      "learning_rate": 0.00018962013431854702,
+      "loss": 0.1294,
+      "step": 2780
+    },
+    {
+      "epoch": 72.94117647058823,
+      "grad_norm": 0.8160730004310608,
+      "learning_rate": 0.00018954665605888088,
+      "loss": 0.1358,
+      "step": 2790
+    },
+    {
+      "epoch": 73.20261437908496,
+      "grad_norm": 0.6369748711585999,
+      "learning_rate": 0.00018947293298207635,
+      "loss": 0.1265,
+      "step": 2800
+    },
+    {
+      "epoch": 73.4640522875817,
+      "grad_norm": 0.8559942245483398,
+      "learning_rate": 0.0001893989652896896,
+      "loss": 0.1301,
+      "step": 2810
+    },
+    {
+      "epoch": 73.72549019607843,
+      "grad_norm": 0.5984659790992737,
+      "learning_rate": 0.0001893247531839454,
+      "loss": 0.1317,
+      "step": 2820
+    },
+    {
+      "epoch": 73.98692810457516,
+      "grad_norm": 0.7031850814819336,
+      "learning_rate": 0.0001892502968677369,
+      "loss": 0.1331,
+      "step": 2830
+    },
+    {
+      "epoch": 74.2483660130719,
+      "grad_norm": 0.7341207265853882,
+      "learning_rate": 0.00018917559654462474,
+      "loss": 0.1268,
+      "step": 2840
+    },
+    {
+      "epoch": 74.50980392156863,
+      "grad_norm": 0.6445427536964417,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.1271,
+      "step": 2850
+    },
+    {
+      "epoch": 74.77124183006536,
+      "grad_norm": 0.7867002487182617,
+      "learning_rate": 0.00018902546469526743,
+      "loss": 0.134,
+      "step": 2860
+    },
+    {
+      "epoch": 75.0326797385621,
+      "grad_norm": 0.5532680153846741,
+      "learning_rate": 0.00018895003357947705,
+      "loss": 0.1271,
+      "step": 2870
+    },
+    {
+      "epoch": 75.29411764705883,
+      "grad_norm": 0.6302035450935364,
+      "learning_rate": 0.00018887435927769137,
+      "loss": 0.1254,
+      "step": 2880
+    },
+    {
+      "epoch": 75.55555555555556,
+      "grad_norm": 0.6897308826446533,
+      "learning_rate": 0.0001887984419968011,
+      "loss": 0.1271,
+      "step": 2890
+    },
+    {
+      "epoch": 75.81699346405229,
+      "grad_norm": 0.5863034725189209,
+      "learning_rate": 0.0001887222819443612,
+      "loss": 0.1258,
+      "step": 2900
+    },
+    {
+      "epoch": 76.07843137254902,
+      "grad_norm": 0.6140713095664978,
+      "learning_rate": 0.00018864587932859028,
+      "loss": 0.1323,
+      "step": 2910
+    },
+    {
+      "epoch": 76.33986928104575,
+      "grad_norm": 0.8416547775268555,
+      "learning_rate": 0.00018856923435837022,
+      "loss": 0.1254,
+      "step": 2920
+    },
+    {
+      "epoch": 76.60130718954248,
+      "grad_norm": 0.6811145544052124,
+      "learning_rate": 0.00018849234724324543,
+      "loss": 0.1269,
+      "step": 2930
+    },
+    {
+      "epoch": 76.86274509803921,
+      "grad_norm": 0.7164523601531982,
+      "learning_rate": 0.00018841521819342236,
+      "loss": 0.1328,
+      "step": 2940
+    },
+    {
+      "epoch": 77.12418300653594,
+      "grad_norm": 0.6024401783943176,
+      "learning_rate": 0.0001883378474197689,
+      "loss": 0.1281,
+      "step": 2950
+    },
+    {
+      "epoch": 77.38562091503267,
+      "grad_norm": 0.6707399487495422,
+      "learning_rate": 0.0001882602351338137,
+      "loss": 0.1278,
+      "step": 2960
+    },
+    {
+      "epoch": 77.6470588235294,
+      "grad_norm": 0.7291834354400635,
+      "learning_rate": 0.00018818238154774588,
+      "loss": 0.1306,
+      "step": 2970
+    },
+    {
+      "epoch": 77.90849673202614,
+      "grad_norm": 0.8068193793296814,
+      "learning_rate": 0.00018810428687441414,
+      "loss": 0.1321,
+      "step": 2980
+    },
+    {
+      "epoch": 78.16993464052288,
+      "grad_norm": 0.5824514627456665,
+      "learning_rate": 0.00018802595132732635,
+      "loss": 0.1228,
+      "step": 2990
+    },
+    {
+      "epoch": 78.43137254901961,
+      "grad_norm": 0.7537580132484436,
+      "learning_rate": 0.0001879473751206489,
+      "loss": 0.1255,
+      "step": 3000
+    },
+    {
+      "epoch": 78.69281045751634,
+      "grad_norm": 0.7544134259223938,
+      "learning_rate": 0.00018786855846920615,
+      "loss": 0.1286,
+      "step": 3010
+    },
+    {
+      "epoch": 78.95424836601308,
+      "grad_norm": 0.695929229259491,
+      "learning_rate": 0.00018778950158847976,
+      "loss": 0.1332,
+      "step": 3020
+    },
+    {
+      "epoch": 79.2156862745098,
+      "grad_norm": 0.5690885782241821,
+      "learning_rate": 0.0001877102046946083,
+      "loss": 0.1209,
+      "step": 3030
+    },
+    {
+      "epoch": 79.47712418300654,
+      "grad_norm": 0.7582102417945862,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.1267,
+      "step": 3040
+    },
+    {
+      "epoch": 79.73856209150327,
+      "grad_norm": 0.48170891404151917,
+      "learning_rate": 0.0001875508917352643,
+      "loss": 0.1304,
+      "step": 3050
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 0.7321044206619263,
+      "learning_rate": 0.00018747087610534736,
+      "loss": 0.1319,
+      "step": 3060
+    },
+    {
+      "epoch": 80.26143790849673,
+      "grad_norm": 0.6163921356201172,
+      "learning_rate": 0.00018739062133339517,
+      "loss": 0.1239,
+      "step": 3070
+    },
+    {
+      "epoch": 80.52287581699346,
+      "grad_norm": 0.8649224042892456,
+      "learning_rate": 0.00018731012763882133,
+      "loss": 0.1294,
+      "step": 3080
+    },
+    {
+      "epoch": 80.7843137254902,
+      "grad_norm": 0.635418713092804,
+      "learning_rate": 0.0001872293952416924,
+      "loss": 0.1274,
+      "step": 3090
+    },
+    {
+      "epoch": 81.04575163398692,
+      "grad_norm": 0.5143495202064514,
+      "learning_rate": 0.00018714842436272773,
+      "loss": 0.1272,
+      "step": 3100
+    },
+    {
+      "epoch": 81.30718954248366,
+      "grad_norm": 0.5938199758529663,
+      "learning_rate": 0.00018706721522329862,
+      "loss": 0.1274,
+      "step": 3110
+    },
+    {
+      "epoch": 81.56862745098039,
+      "grad_norm": 0.6700434684753418,
+      "learning_rate": 0.00018698576804542777,
+      "loss": 0.1239,
+      "step": 3120
+    },
+    {
+      "epoch": 81.83006535947712,
+      "grad_norm": 0.6935993432998657,
+      "learning_rate": 0.00018690408305178856,
+      "loss": 0.129,
+      "step": 3130
+    },
+    {
+      "epoch": 82.09150326797386,
+      "grad_norm": 0.616468071937561,
+      "learning_rate": 0.00018682216046570475,
+      "loss": 0.1309,
+      "step": 3140
+    },
+    {
+      "epoch": 82.3529411764706,
+      "grad_norm": 0.6784673929214478,
+      "learning_rate": 0.00018674000051114952,
+      "loss": 0.1256,
+      "step": 3150
+    },
+    {
+      "epoch": 82.61437908496733,
+      "grad_norm": 0.6315344572067261,
+      "learning_rate": 0.00018665760341274505,
+      "loss": 0.1248,
+      "step": 3160
+    },
+    {
+      "epoch": 82.87581699346406,
+      "grad_norm": 0.5992943644523621,
+      "learning_rate": 0.00018657496939576186,
+      "loss": 0.1289,
+      "step": 3170
+    },
+    {
+      "epoch": 83.13725490196079,
+      "grad_norm": 0.5748761892318726,
+      "learning_rate": 0.0001864920986861182,
+      "loss": 0.13,
+      "step": 3180
+    },
+    {
+      "epoch": 83.39869281045752,
+      "grad_norm": 0.4966670572757721,
+      "learning_rate": 0.00018640899151037944,
+      "loss": 0.125,
+      "step": 3190
+    },
+    {
+      "epoch": 83.66013071895425,
+      "grad_norm": 0.7322400212287903,
+      "learning_rate": 0.00018632564809575742,
+      "loss": 0.1296,
+      "step": 3200
+    },
+    {
+      "epoch": 83.92156862745098,
+      "grad_norm": 0.8756809234619141,
+      "learning_rate": 0.0001862420686701098,
+      "loss": 0.1258,
+      "step": 3210
+    },
+    {
+      "epoch": 84.18300653594771,
+      "grad_norm": 0.6855464577674866,
+      "learning_rate": 0.0001861582534619396,
+      "loss": 0.1293,
+      "step": 3220
+    },
+    {
+      "epoch": 84.44444444444444,
+      "grad_norm": 0.5204270482063293,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 0.1239,
+      "step": 3230
+    },
+    {
+      "epoch": 84.70588235294117,
+      "grad_norm": 0.6892998814582825,
+      "learning_rate": 0.00018598991661526572,
+      "loss": 0.1276,
+      "step": 3240
+    },
+    {
+      "epoch": 84.9673202614379,
+      "grad_norm": 0.5613439083099365,
+      "learning_rate": 0.00018590539543698854,
+      "loss": 0.1291,
+      "step": 3250
+    },
+    {
+      "epoch": 85.22875816993464,
+      "grad_norm": 0.6739300489425659,
+      "learning_rate": 0.0001858206393966405,
+      "loss": 0.1264,
+      "step": 3260
+    },
+    {
+      "epoch": 85.49019607843137,
+      "grad_norm": 0.6257854700088501,
+      "learning_rate": 0.00018573564872594145,
+      "loss": 0.1228,
+      "step": 3270
+    },
+    {
+      "epoch": 85.7516339869281,
+      "grad_norm": 0.7718020081520081,
+      "learning_rate": 0.00018565042365725258,
+      "loss": 0.1272,
+      "step": 3280
+    },
+    {
+      "epoch": 86.01307189542484,
+      "grad_norm": 0.599905252456665,
+      "learning_rate": 0.00018556496442357595,
+      "loss": 0.131,
+      "step": 3290
+    },
+    {
+      "epoch": 86.27450980392157,
+      "grad_norm": 0.4919801950454712,
+      "learning_rate": 0.0001854792712585539,
+      "loss": 0.1187,
+      "step": 3300
+    },
+    {
+      "epoch": 86.5359477124183,
+      "grad_norm": 0.7735366821289062,
+      "learning_rate": 0.00018539334439646824,
+      "loss": 0.1292,
+      "step": 3310
+    },
+    {
+      "epoch": 86.79738562091504,
+      "grad_norm": 0.5821804404258728,
+      "learning_rate": 0.00018530718407223974,
+      "loss": 0.1241,
+      "step": 3320
+    },
+    {
+      "epoch": 87.05882352941177,
+      "grad_norm": 0.5197769403457642,
+      "learning_rate": 0.00018522079052142747,
+      "loss": 0.1327,
+      "step": 3330
+    },
+    {
+      "epoch": 87.3202614379085,
+      "grad_norm": 0.728046715259552,
+      "learning_rate": 0.00018513416398022802,
+      "loss": 0.125,
+      "step": 3340
+    },
+    {
+      "epoch": 87.58169934640523,
+      "grad_norm": 0.5133826732635498,
+      "learning_rate": 0.0001850473046854751,
+      "loss": 0.125,
+      "step": 3350
+    },
+    {
+      "epoch": 87.84313725490196,
+      "grad_norm": 0.5190628170967102,
+      "learning_rate": 0.0001849602128746387,
+      "loss": 0.1266,
+      "step": 3360
+    },
+    {
+      "epoch": 88.10457516339869,
+      "grad_norm": 0.5673719048500061,
+      "learning_rate": 0.00018487288878582447,
+      "loss": 0.1264,
+      "step": 3370
+    },
+    {
+      "epoch": 88.36601307189542,
+      "grad_norm": 0.5447489023208618,
+      "learning_rate": 0.00018478533265777318,
+      "loss": 0.1227,
+      "step": 3380
+    },
+    {
+      "epoch": 88.62745098039215,
+      "grad_norm": 0.7577037811279297,
+      "learning_rate": 0.00018469754472985993,
+      "loss": 0.1301,
+      "step": 3390
+    },
+    {
+      "epoch": 88.88888888888889,
+      "grad_norm": 0.6202467083930969,
+      "learning_rate": 0.00018460952524209355,
+      "loss": 0.1282,
+      "step": 3400
+    },
+    {
+      "epoch": 89.15032679738562,
+      "grad_norm": 0.5380336046218872,
+      "learning_rate": 0.00018452127443511598,
+      "loss": 0.1208,
+      "step": 3410
+    },
+    {
+      "epoch": 89.41176470588235,
+      "grad_norm": 0.6075873970985413,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 0.1253,
+      "step": 3420
+    },
+    {
+      "epoch": 89.67320261437908,
+      "grad_norm": 0.6594501733779907,
+      "learning_rate": 0.0001843440798292563,
+      "loss": 0.1292,
+      "step": 3430
+    },
+    {
+      "epoch": 89.93464052287581,
+      "grad_norm": 0.8450397849082947,
+      "learning_rate": 0.00018425513651481747,
+      "loss": 0.1272,
+      "step": 3440
+    },
+    {
+      "epoch": 90.19607843137256,
+      "grad_norm": 0.5860391855239868,
+      "learning_rate": 0.00018416596285005272,
+      "loss": 0.1235,
+      "step": 3450
+    },
+    {
+      "epoch": 90.45751633986929,
+      "grad_norm": 1.0777153968811035,
+      "learning_rate": 0.0001840765590787594,
+      "loss": 0.1236,
+      "step": 3460
+    },
+    {
+      "epoch": 90.71895424836602,
+      "grad_norm": 0.5201640129089355,
+      "learning_rate": 0.00018398692544536397,
+      "loss": 0.1257,
+      "step": 3470
+    },
+    {
+      "epoch": 90.98039215686275,
+      "grad_norm": 0.6437315344810486,
+      "learning_rate": 0.00018389706219492147,
+      "loss": 0.1309,
+      "step": 3480
+    },
+    {
+      "epoch": 91.24183006535948,
+      "grad_norm": 0.7315922975540161,
+      "learning_rate": 0.00018380696957311449,
+      "loss": 0.1202,
+      "step": 3490
+    },
+    {
+      "epoch": 91.50326797385621,
+      "grad_norm": 0.6596841216087341,
+      "learning_rate": 0.00018371664782625287,
+      "loss": 0.123,
+      "step": 3500
+    },
+    {
+      "epoch": 91.76470588235294,
+      "grad_norm": 0.7001237869262695,
+      "learning_rate": 0.0001836260972012728,
+      "loss": 0.131,
+      "step": 3510
+    },
+    {
+      "epoch": 92.02614379084967,
+      "grad_norm": 0.5139759182929993,
+      "learning_rate": 0.00018353531794573625,
+      "loss": 0.1269,
+      "step": 3520
+    },
+    {
+      "epoch": 92.2875816993464,
+      "grad_norm": 0.5682742595672607,
+      "learning_rate": 0.0001834443103078302,
+      "loss": 0.1219,
+      "step": 3530
+    },
+    {
+      "epoch": 92.54901960784314,
+      "grad_norm": 0.5991658568382263,
+      "learning_rate": 0.0001833530745363661,
+      "loss": 0.1253,
+      "step": 3540
+    },
+    {
+      "epoch": 92.81045751633987,
+      "grad_norm": 0.600954532623291,
+      "learning_rate": 0.00018326161088077903,
+      "loss": 0.1286,
+      "step": 3550
+    },
+    {
+      "epoch": 93.0718954248366,
+      "grad_norm": 0.6942703723907471,
+      "learning_rate": 0.00018316991959112716,
+      "loss": 0.129,
+      "step": 3560
+    },
+    {
+      "epoch": 93.33333333333333,
+      "grad_norm": 0.723839282989502,
+      "learning_rate": 0.00018307800091809097,
+      "loss": 0.1243,
+      "step": 3570
+    },
+    {
+      "epoch": 93.59477124183006,
+      "grad_norm": 0.8791704773902893,
+      "learning_rate": 0.0001829858551129726,
+      "loss": 0.1266,
+      "step": 3580
+    },
+    {
+      "epoch": 93.85620915032679,
+      "grad_norm": 0.6914976835250854,
+      "learning_rate": 0.00018289348242769515,
+      "loss": 0.1252,
+      "step": 3590
+    },
+    {
+      "epoch": 94.11764705882354,
+      "grad_norm": 0.40806615352630615,
+      "learning_rate": 0.00018280088311480201,
+      "loss": 0.1245,
+      "step": 3600
+    },
+    {
+      "epoch": 94.37908496732027,
+      "grad_norm": 0.47459840774536133,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 0.124,
+      "step": 3610
+    },
+    {
+      "epoch": 94.640522875817,
+      "grad_norm": 0.5964429378509521,
+      "learning_rate": 0.00018261500561943955,
+      "loss": 0.1247,
+      "step": 3620
+    },
+    {
+      "epoch": 94.90196078431373,
+      "grad_norm": 0.5425547957420349,
+      "learning_rate": 0.00018252172794515223,
+      "loss": 0.1291,
+      "step": 3630
+    },
+    {
+      "epoch": 95.16339869281046,
+      "grad_norm": 0.6276856660842896,
+      "learning_rate": 0.00018242822465961176,
+      "loss": 0.1291,
+      "step": 3640
+    },
+    {
+      "epoch": 95.42483660130719,
+      "grad_norm": 1.3327772617340088,
+      "learning_rate": 0.00018233449601845258,
+      "loss": 0.1209,
+      "step": 3650
+    },
+    {
+      "epoch": 95.68627450980392,
+      "grad_norm": 0.9959401488304138,
+      "learning_rate": 0.00018224054227792524,
+      "loss": 0.1237,
+      "step": 3660
+    },
+    {
+      "epoch": 95.94771241830065,
+      "grad_norm": 0.6117873787879944,
+      "learning_rate": 0.00018214636369489563,
+      "loss": 0.1336,
+      "step": 3670
+    },
+    {
+      "epoch": 96.20915032679738,
+      "grad_norm": 0.5268058776855469,
+      "learning_rate": 0.00018205196052684445,
+      "loss": 0.1212,
+      "step": 3680
+    },
+    {
+      "epoch": 96.47058823529412,
+      "grad_norm": 0.6816233396530151,
+      "learning_rate": 0.00018195733303186633,
+      "loss": 0.1291,
+      "step": 3690
+    },
+    {
+      "epoch": 96.73202614379085,
+      "grad_norm": 0.6062767505645752,
+      "learning_rate": 0.00018186248146866927,
+      "loss": 0.1248,
+      "step": 3700
+    },
+    {
+      "epoch": 96.99346405228758,
+      "grad_norm": 0.9882863759994507,
+      "learning_rate": 0.0001817674060965737,
+      "loss": 0.1297,
+      "step": 3710
+    },
+    {
+      "epoch": 97.25490196078431,
+      "grad_norm": 0.5438330769538879,
+      "learning_rate": 0.00018167210717551224,
+      "loss": 0.1207,
+      "step": 3720
+    },
+    {
+      "epoch": 97.51633986928104,
+      "grad_norm": 0.5340378284454346,
+      "learning_rate": 0.00018157658496602833,
+      "loss": 0.1219,
+      "step": 3730
+    },
+    {
+      "epoch": 97.77777777777777,
+      "grad_norm": 0.6972679495811462,
+      "learning_rate": 0.00018148083972927616,
+      "loss": 0.1272,
+      "step": 3740
+    },
+    {
+      "epoch": 98.03921568627452,
+      "grad_norm": 0.5716882348060608,
+      "learning_rate": 0.0001813848717270195,
+      "loss": 0.1293,
+      "step": 3750
+    },
+    {
+      "epoch": 98.30065359477125,
+      "grad_norm": 0.6160675883293152,
+      "learning_rate": 0.00018128868122163123,
+      "loss": 0.1231,
+      "step": 3760
+    },
+    {
+      "epoch": 98.56209150326798,
+      "grad_norm": 0.6029688715934753,
+      "learning_rate": 0.00018119226847609245,
+      "loss": 0.1243,
+      "step": 3770
+    },
+    {
+      "epoch": 98.82352941176471,
+      "grad_norm": 0.48037102818489075,
+      "learning_rate": 0.000181095633753992,
+      "loss": 0.1244,
+      "step": 3780
+    },
+    {
+      "epoch": 99.08496732026144,
+      "grad_norm": 0.5181258916854858,
+      "learning_rate": 0.0001809987773195255,
+      "loss": 0.1293,
+      "step": 3790
+    },
+    {
+      "epoch": 99.34640522875817,
+      "grad_norm": 0.5500502586364746,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.1223,
+      "step": 3800
+    },
+    {
+      "epoch": 99.6078431372549,
+      "grad_norm": 0.5222991704940796,
+      "learning_rate": 0.00018080440037330695,
+      "loss": 0.1234,
+      "step": 3810
+    },
+    {
+      "epoch": 99.86928104575163,
+      "grad_norm": 0.44636550545692444,
+      "learning_rate": 0.00018070688039297403,
+      "loss": 0.1229,
+      "step": 3820
+    },
+    {
+      "epoch": 100.13071895424837,
+      "grad_norm": 0.6223607063293457,
+      "learning_rate": 0.00018060913976311192,
+      "loss": 0.1248,
+      "step": 3830
+    },
+    {
+      "epoch": 100.3921568627451,
+      "grad_norm": 0.5242628455162048,
+      "learning_rate": 0.00018051117875093976,
+      "loss": 0.1228,
+      "step": 3840
+    },
+    {
+      "epoch": 100.65359477124183,
+      "grad_norm": 0.5237357020378113,
+      "learning_rate": 0.00018041299762427916,
+      "loss": 0.1249,
+      "step": 3850
+    },
+    {
+      "epoch": 100.91503267973856,
+      "grad_norm": 0.48059752583503723,
+      "learning_rate": 0.00018031459665155363,
+      "loss": 0.1244,
+      "step": 3860
+    },
+    {
+      "epoch": 101.17647058823529,
+      "grad_norm": 0.6612676382064819,
+      "learning_rate": 0.00018021597610178768,
+      "loss": 0.1232,
+      "step": 3870
+    },
+    {
+      "epoch": 101.43790849673202,
+      "grad_norm": 0.5979042053222656,
+      "learning_rate": 0.00018011713624460608,
+      "loss": 0.1238,
+      "step": 3880
+    },
+    {
+      "epoch": 101.69934640522875,
+      "grad_norm": 0.5387808084487915,
+      "learning_rate": 0.0001800180773502333,
+      "loss": 0.1256,
+      "step": 3890
+    },
+    {
+      "epoch": 101.96078431372548,
+      "grad_norm": 0.526494026184082,
+      "learning_rate": 0.0001799187996894925,
+      "loss": 0.1237,
+      "step": 3900
+    },
+    {
+      "epoch": 102.22222222222223,
+      "grad_norm": 0.5353228449821472,
+      "learning_rate": 0.00017981930353380503,
+      "loss": 0.1196,
+      "step": 3910
+    },
+    {
+      "epoch": 102.48366013071896,
+      "grad_norm": 0.4083467125892639,
+      "learning_rate": 0.0001797195891551896,
+      "loss": 0.1242,
+      "step": 3920
+    },
+    {
+      "epoch": 102.74509803921569,
+      "grad_norm": 0.4955314099788666,
+      "learning_rate": 0.0001796196568262615,
+      "loss": 0.1273,
+      "step": 3930
+    },
+    {
+      "epoch": 103.00653594771242,
+      "grad_norm": 0.6663371920585632,
+      "learning_rate": 0.00017951950682023191,
+      "loss": 0.1271,
+      "step": 3940
+    },
+    {
+      "epoch": 103.26797385620915,
+      "grad_norm": 0.5380375385284424,
+      "learning_rate": 0.0001794191394109071,
+      "loss": 0.1197,
+      "step": 3950
+    },
+    {
+      "epoch": 103.52941176470588,
+      "grad_norm": 0.5058585405349731,
+      "learning_rate": 0.00017931855487268782,
+      "loss": 0.1253,
+      "step": 3960
+    },
+    {
+      "epoch": 103.79084967320262,
+      "grad_norm": 0.4719361364841461,
+      "learning_rate": 0.0001792177534805682,
+      "loss": 0.1221,
+      "step": 3970
+    },
+    {
+      "epoch": 104.05228758169935,
+      "grad_norm": 0.5762635469436646,
+      "learning_rate": 0.00017911673551013551,
+      "loss": 0.1273,
+      "step": 3980
+    },
+    {
+      "epoch": 104.31372549019608,
+      "grad_norm": 0.5460364818572998,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 0.1207,
+      "step": 3990
+    },
+    {
+      "epoch": 104.57516339869281,
+      "grad_norm": 0.6880167722702026,
+      "learning_rate": 0.00017891405093963938,
+      "loss": 0.123,
+      "step": 4000
+    },
+    {
+      "epoch": 104.83660130718954,
+      "grad_norm": 0.5890039205551147,
+      "learning_rate": 0.00017881238489370776,
+      "loss": 0.1261,
+      "step": 4010
+    },
+    {
+      "epoch": 105.09803921568627,
+      "grad_norm": 0.5330226421356201,
+      "learning_rate": 0.00017871050337772525,
+      "loss": 0.1257,
+      "step": 4020
+    },
+    {
+      "epoch": 105.359477124183,
+      "grad_norm": 0.7687032222747803,
+      "learning_rate": 0.00017860840667023212,
+      "loss": 0.1191,
+      "step": 4030
+    },
+    {
+      "epoch": 105.62091503267973,
+      "grad_norm": 0.5523906946182251,
+      "learning_rate": 0.0001785060950503568,
+      "loss": 0.1222,
+      "step": 4040
+    },
+    {
+      "epoch": 105.88235294117646,
+      "grad_norm": 0.710148811340332,
+      "learning_rate": 0.0001784035687978153,
+      "loss": 0.1285,
+      "step": 4050
+    },
+    {
+      "epoch": 106.14379084967321,
+      "grad_norm": 0.5394968390464783,
+      "learning_rate": 0.0001783008281929106,
+      "loss": 0.1236,
+      "step": 4060
+    },
+    {
+      "epoch": 106.40522875816994,
+      "grad_norm": 0.9128403663635254,
+      "learning_rate": 0.0001781978735165315,
+      "loss": 0.123,
+      "step": 4070
+    },
+    {
+      "epoch": 106.66666666666667,
+      "grad_norm": 0.6581969857215881,
+      "learning_rate": 0.0001780947050501522,
+      "loss": 0.1233,
+      "step": 4080
+    },
+    {
+      "epoch": 106.9281045751634,
+      "grad_norm": 0.7595931887626648,
+      "learning_rate": 0.00017799132307583132,
+      "loss": 0.1265,
+      "step": 4090
+    },
+    {
+      "epoch": 107.18954248366013,
+      "grad_norm": 0.4747548997402191,
+      "learning_rate": 0.00017788772787621126,
+      "loss": 0.1205,
+      "step": 4100
+    },
+    {
+      "epoch": 107.45098039215686,
+      "grad_norm": 0.6191473603248596,
+      "learning_rate": 0.0001777839197345173,
+      "loss": 0.1227,
+      "step": 4110
+    },
+    {
+      "epoch": 107.7124183006536,
+      "grad_norm": 0.6291956305503845,
+      "learning_rate": 0.00017767989893455698,
+      "loss": 0.1241,
+      "step": 4120
+    },
+    {
+      "epoch": 107.97385620915033,
+      "grad_norm": 0.8078779578208923,
+      "learning_rate": 0.00017757566576071914,
+      "loss": 0.1276,
+      "step": 4130
+    },
+    {
+      "epoch": 108.23529411764706,
+      "grad_norm": 0.534305989742279,
+      "learning_rate": 0.00017747122049797335,
+      "loss": 0.1237,
+      "step": 4140
+    },
+    {
+      "epoch": 108.49673202614379,
+      "grad_norm": 0.6715922951698303,
+      "learning_rate": 0.00017736656343186896,
+      "loss": 0.1222,
+      "step": 4150
+    },
+    {
+      "epoch": 108.75816993464052,
+      "grad_norm": 0.4797965884208679,
+      "learning_rate": 0.00017726169484853438,
+      "loss": 0.1233,
+      "step": 4160
+    },
+    {
+      "epoch": 109.01960784313725,
+      "grad_norm": 0.5531324148178101,
+      "learning_rate": 0.0001771566150346763,
+      "loss": 0.1248,
+      "step": 4170
+    },
+    {
+      "epoch": 109.28104575163398,
+      "grad_norm": 0.6008434891700745,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 0.1191,
+      "step": 4180
+    },
+    {
+      "epoch": 109.54248366013071,
+      "grad_norm": 0.4768673777580261,
+      "learning_rate": 0.0001769458228651032,
+      "loss": 0.1221,
+      "step": 4190
+    },
+    {
+      "epoch": 109.80392156862744,
+      "grad_norm": 0.6039496064186096,
+      "learning_rate": 0.00017684011108568592,
+      "loss": 0.1272,
+      "step": 4200
+    },
+    {
+      "epoch": 110.06535947712419,
+      "grad_norm": 0.4699447453022003,
+      "learning_rate": 0.00017673418922833903,
+      "loss": 0.124,
+      "step": 4210
+    },
+    {
+      "epoch": 110.32679738562092,
+      "grad_norm": 0.5179737210273743,
+      "learning_rate": 0.00017662805758264893,
+      "loss": 0.1216,
+      "step": 4220
+    },
+    {
+      "epoch": 110.58823529411765,
+      "grad_norm": 0.5560812950134277,
+      "learning_rate": 0.0001765217164387754,
+      "loss": 0.1254,
+      "step": 4230
+    },
+    {
+      "epoch": 110.84967320261438,
+      "grad_norm": 0.5947309732437134,
+      "learning_rate": 0.00017641516608745114,
+      "loss": 0.126,
+      "step": 4240
+    },
+    {
+      "epoch": 111.11111111111111,
+      "grad_norm": 0.5091108083724976,
+      "learning_rate": 0.00017630840681998066,
+      "loss": 0.1253,
+      "step": 4250
+    },
+    {
+      "epoch": 111.37254901960785,
+      "grad_norm": 0.6985629796981812,
+      "learning_rate": 0.00017620143892823977,
+      "loss": 0.1234,
+      "step": 4260
+    },
+    {
+      "epoch": 111.63398692810458,
+      "grad_norm": 0.4949619770050049,
+      "learning_rate": 0.00017609426270467462,
+      "loss": 0.1243,
+      "step": 4270
+    },
+    {
+      "epoch": 111.89542483660131,
+      "grad_norm": 0.6455708742141724,
+      "learning_rate": 0.00017598687844230088,
+      "loss": 0.1268,
+      "step": 4280
+    },
+    {
+      "epoch": 112.15686274509804,
+      "grad_norm": 0.45366206765174866,
+      "learning_rate": 0.000175879286434703,
+      "loss": 0.1218,
+      "step": 4290
+    },
+    {
+      "epoch": 112.41830065359477,
+      "grad_norm": 0.4446161687374115,
+      "learning_rate": 0.0001757714869760335,
+      "loss": 0.1159,
+      "step": 4300
+    },
+    {
+      "epoch": 112.6797385620915,
+      "grad_norm": 0.5011366009712219,
+      "learning_rate": 0.00017566348036101187,
+      "loss": 0.1237,
+      "step": 4310
+    },
+    {
+      "epoch": 112.94117647058823,
+      "grad_norm": 0.6293872594833374,
+      "learning_rate": 0.0001755552668849242,
+      "loss": 0.1261,
+      "step": 4320
+    },
+    {
+      "epoch": 113.20261437908496,
+      "grad_norm": 0.5598844885826111,
+      "learning_rate": 0.0001754468468436219,
+      "loss": 0.1257,
+      "step": 4330
+    },
+    {
+      "epoch": 113.4640522875817,
+      "grad_norm": 0.4953804314136505,
+      "learning_rate": 0.00017533822053352128,
+      "loss": 0.1193,
+      "step": 4340
+    },
+    {
+      "epoch": 113.72549019607843,
+      "grad_norm": 0.4813307225704193,
+      "learning_rate": 0.0001752293882516025,
+      "loss": 0.1203,
+      "step": 4350
+    },
+    {
+      "epoch": 113.98692810457516,
+      "grad_norm": 0.5343033075332642,
+      "learning_rate": 0.00017512035029540885,
+      "loss": 0.1287,
+      "step": 4360
+    },
+    {
+      "epoch": 114.2483660130719,
+      "grad_norm": 0.3980255722999573,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 0.1194,
+      "step": 4370
+    },
+    {
+      "epoch": 114.50980392156863,
+      "grad_norm": 0.49828794598579407,
+      "learning_rate": 0.00017490165855318094,
+      "loss": 0.1225,
+      "step": 4380
+    },
+    {
+      "epoch": 114.77124183006536,
+      "grad_norm": 0.5137373208999634,
+      "learning_rate": 0.00017479200536504156,
+      "loss": 0.1246,
+      "step": 4390
+    },
+    {
+      "epoch": 115.0326797385621,
+      "grad_norm": 0.4799012839794159,
+      "learning_rate": 0.0001746821476984154,
+      "loss": 0.1267,
+      "step": 4400
+    },
+    {
+      "epoch": 115.29411764705883,
+      "grad_norm": 0.4675128757953644,
+      "learning_rate": 0.00017457208585364918,
+      "loss": 0.1212,
+      "step": 4410
+    },
+    {
+      "epoch": 115.55555555555556,
+      "grad_norm": 0.5032903552055359,
+      "learning_rate": 0.00017446182013164778,
+      "loss": 0.1233,
+      "step": 4420
+    },
+    {
+      "epoch": 115.81699346405229,
+      "grad_norm": 0.5266563892364502,
+      "learning_rate": 0.00017435135083387345,
+      "loss": 0.1259,
+      "step": 4430
+    },
+    {
+      "epoch": 116.07843137254902,
+      "grad_norm": 0.5633729696273804,
+      "learning_rate": 0.000174240678262345,
+      "loss": 0.1229,
+      "step": 4440
+    },
+    {
+      "epoch": 116.33986928104575,
+      "grad_norm": 0.6905789375305176,
+      "learning_rate": 0.0001741298027196371,
+      "loss": 0.1198,
+      "step": 4450
+    },
+    {
+      "epoch": 116.60130718954248,
+      "grad_norm": 0.39248159527778625,
+      "learning_rate": 0.00017401872450887917,
+      "loss": 0.1201,
+      "step": 4460
+    },
+    {
+      "epoch": 116.86274509803921,
+      "grad_norm": 0.48288437724113464,
+      "learning_rate": 0.00017390744393375486,
+      "loss": 0.1258,
+      "step": 4470
+    },
+    {
+      "epoch": 117.12418300653594,
+      "grad_norm": 0.5383310914039612,
+      "learning_rate": 0.00017379596129850098,
+      "loss": 0.1219,
+      "step": 4480
+    },
+    {
+      "epoch": 117.38562091503267,
+      "grad_norm": 0.5818719267845154,
+      "learning_rate": 0.00017368427690790677,
+      "loss": 0.122,
+      "step": 4490
+    },
+    {
+      "epoch": 117.6470588235294,
+      "grad_norm": 0.6253119111061096,
+      "learning_rate": 0.00017357239106731317,
+      "loss": 0.1218,
+      "step": 4500
+    },
+    {
+      "epoch": 117.90849673202614,
+      "grad_norm": 0.560431182384491,
+      "learning_rate": 0.00017346030408261172,
+      "loss": 0.1254,
+      "step": 4510
+    },
+    {
+      "epoch": 118.16993464052288,
+      "grad_norm": 0.4778011441230774,
+      "learning_rate": 0.000173348016260244,
+      "loss": 0.1264,
+      "step": 4520
+    },
+    {
+      "epoch": 118.43137254901961,
+      "grad_norm": 0.44935017824172974,
+      "learning_rate": 0.00017323552790720058,
+      "loss": 0.1194,
+      "step": 4530
+    },
+    {
+      "epoch": 118.69281045751634,
+      "grad_norm": 0.5576562881469727,
+      "learning_rate": 0.00017312283933102038,
+      "loss": 0.1262,
+      "step": 4540
+    },
+    {
+      "epoch": 118.95424836601308,
+      "grad_norm": 0.43395355343818665,
+      "learning_rate": 0.00017300995083978965,
+      "loss": 0.1206,
+      "step": 4550
+    },
+    {
+      "epoch": 119.2156862745098,
+      "grad_norm": 0.5254650712013245,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.1222,
+      "step": 4560
+    },
+    {
+      "epoch": 119.47712418300654,
+      "grad_norm": 0.5673931241035461,
+      "learning_rate": 0.0001727835753472535,
+      "loss": 0.1186,
+      "step": 4570
+    },
+    {
+      "epoch": 119.73856209150327,
+      "grad_norm": 0.38074398040771484,
+      "learning_rate": 0.0001726700889648501,
+      "loss": 0.1214,
+      "step": 4580
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 0.5379620790481567,
+      "learning_rate": 0.00017255640390519836,
+      "loss": 0.1262,
+      "step": 4590
+    },
+    {
+      "epoch": 120.26143790849673,
+      "grad_norm": 0.46410471200942993,
+      "learning_rate": 0.00017244252047910892,
+      "loss": 0.1187,
+      "step": 4600
+    },
+    {
+      "epoch": 120.52287581699346,
+      "grad_norm": 0.5001809000968933,
+      "learning_rate": 0.00017232843899793468,
+      "loss": 0.1183,
+      "step": 4610
+    },
+    {
+      "epoch": 120.7843137254902,
+      "grad_norm": 0.402111291885376,
+      "learning_rate": 0.00017221415977357007,
+      "loss": 0.1243,
+      "step": 4620
+    },
+    {
+      "epoch": 121.04575163398692,
+      "grad_norm": 0.389432817697525,
+      "learning_rate": 0.00017209968311845012,
+      "loss": 0.1246,
+      "step": 4630
+    },
+    {
+      "epoch": 121.30718954248366,
+      "grad_norm": 0.5195785760879517,
+      "learning_rate": 0.00017198500934554966,
+      "loss": 0.1171,
+      "step": 4640
+    },
+    {
+      "epoch": 121.56862745098039,
+      "grad_norm": 0.4785175323486328,
+      "learning_rate": 0.0001718701387683824,
+      "loss": 0.1199,
+      "step": 4650
+    },
+    {
+      "epoch": 121.83006535947712,
+      "grad_norm": 0.5644505620002747,
+      "learning_rate": 0.0001717550717010001,
+      "loss": 0.1251,
+      "step": 4660
+    },
+    {
+      "epoch": 122.09150326797386,
+      "grad_norm": 0.506691038608551,
+      "learning_rate": 0.0001716398084579917,
+      "loss": 0.1259,
+      "step": 4670
+    },
+    {
+      "epoch": 122.3529411764706,
+      "grad_norm": 0.491377592086792,
+      "learning_rate": 0.00017152434935448256,
+      "loss": 0.1203,
+      "step": 4680
+    },
+    {
+      "epoch": 122.61437908496733,
+      "grad_norm": 0.5997154712677002,
+      "learning_rate": 0.00017140869470613342,
+      "loss": 0.1179,
+      "step": 4690
+    },
+    {
+      "epoch": 122.87581699346406,
+      "grad_norm": 0.4591546356678009,
+      "learning_rate": 0.00017129284482913972,
+      "loss": 0.1218,
+      "step": 4700
+    },
+    {
+      "epoch": 123.13725490196079,
+      "grad_norm": 0.3804793655872345,
+      "learning_rate": 0.00017117680004023056,
+      "loss": 0.1195,
+      "step": 4710
+    },
+    {
+      "epoch": 123.39869281045752,
+      "grad_norm": 0.4563983082771301,
+      "learning_rate": 0.00017106056065666793,
+      "loss": 0.1177,
+      "step": 4720
+    },
+    {
+      "epoch": 123.66013071895425,
+      "grad_norm": 0.5122204422950745,
+      "learning_rate": 0.00017094412699624595,
+      "loss": 0.1221,
+      "step": 4730
+    },
+    {
+      "epoch": 123.92156862745098,
+      "grad_norm": 0.4884132146835327,
+      "learning_rate": 0.00017082749937728973,
+      "loss": 0.1245,
+      "step": 4740
+    },
+    {
+      "epoch": 124.18300653594771,
+      "grad_norm": 0.38031864166259766,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.1196,
+      "step": 4750
+    },
+    {
+      "epoch": 124.44444444444444,
+      "grad_norm": 0.4657631814479828,
+      "learning_rate": 0.0001705936635397259,
+      "loss": 0.1197,
+      "step": 4760
+    },
+    {
+      "epoch": 124.70588235294117,
+      "grad_norm": 0.5115662217140198,
+      "learning_rate": 0.00017047645596041653,
+      "loss": 0.1217,
+      "step": 4770
+    },
+    {
+      "epoch": 124.9673202614379,
+      "grad_norm": 0.39813730120658875,
+      "learning_rate": 0.0001703590557011677,
+      "loss": 0.12,
+      "step": 4780
+    },
+    {
+      "epoch": 125.22875816993464,
+      "grad_norm": 0.4247078001499176,
+      "learning_rate": 0.00017024146308294724,
+      "loss": 0.1169,
+      "step": 4790
+    },
+    {
+      "epoch": 125.49019607843137,
+      "grad_norm": 0.41345420479774475,
+      "learning_rate": 0.00017012367842724887,
+      "loss": 0.1212,
+      "step": 4800
+    },
+    {
+      "epoch": 125.7516339869281,
+      "grad_norm": 0.43008559942245483,
+      "learning_rate": 0.00017000570205609136,
+      "loss": 0.1198,
+      "step": 4810
+    },
+    {
+      "epoch": 126.01307189542484,
+      "grad_norm": 0.47986850142478943,
+      "learning_rate": 0.00016988753429201755,
+      "loss": 0.125,
+      "step": 4820
+    },
+    {
+      "epoch": 126.27450980392157,
+      "grad_norm": 0.3475463390350342,
+      "learning_rate": 0.00016976917545809367,
+      "loss": 0.1146,
+      "step": 4830
+    },
+    {
+      "epoch": 126.5359477124183,
+      "grad_norm": 0.38883545994758606,
+      "learning_rate": 0.00016965062587790823,
+      "loss": 0.1189,
+      "step": 4840
+    },
+    {
+      "epoch": 126.79738562091504,
+      "grad_norm": 0.44550979137420654,
+      "learning_rate": 0.00016953188587557122,
+      "loss": 0.1247,
+      "step": 4850
+    },
+    {
+      "epoch": 127.05882352941177,
+      "grad_norm": 0.3760294020175934,
+      "learning_rate": 0.0001694129557757133,
+      "loss": 0.125,
+      "step": 4860
+    },
+    {
+      "epoch": 127.3202614379085,
+      "grad_norm": 0.46050703525543213,
+      "learning_rate": 0.0001692938359034848,
+      "loss": 0.1161,
+      "step": 4870
+    },
+    {
+      "epoch": 127.58169934640523,
+      "grad_norm": 0.48216715455055237,
+      "learning_rate": 0.00016917452658455495,
+      "loss": 0.1198,
+      "step": 4880
+    },
+    {
+      "epoch": 127.84313725490196,
+      "grad_norm": 0.4969286620616913,
+      "learning_rate": 0.00016905502814511082,
+      "loss": 0.1231,
+      "step": 4890
+    },
+    {
+      "epoch": 128.1045751633987,
+      "grad_norm": 0.3721112012863159,
+      "learning_rate": 0.0001689353409118566,
+      "loss": 0.1275,
+      "step": 4900
+    },
+    {
+      "epoch": 128.36601307189542,
+      "grad_norm": 0.518031895160675,
+      "learning_rate": 0.0001688154652120126,
+      "loss": 0.118,
+      "step": 4910
+    },
+    {
+      "epoch": 128.62745098039215,
+      "grad_norm": 0.49686360359191895,
+      "learning_rate": 0.00016869540137331445,
+      "loss": 0.1179,
+      "step": 4920
+    },
+    {
+      "epoch": 128.88888888888889,
+      "grad_norm": 0.5079768300056458,
+      "learning_rate": 0.00016857514972401207,
+      "loss": 0.1252,
+      "step": 4930
+    },
+    {
+      "epoch": 129.15032679738562,
+      "grad_norm": 0.4462049901485443,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 0.1196,
+      "step": 4940
+    },
+    {
+      "epoch": 129.41176470588235,
+      "grad_norm": 0.4287566840648651,
+      "learning_rate": 0.00016833408430916085,
+      "loss": 0.1198,
+      "step": 4950
+    },
+    {
+      "epoch": 129.67320261437908,
+      "grad_norm": 0.4560422897338867,
+      "learning_rate": 0.00016821327120267567,
+      "loss": 0.1218,
+      "step": 4960
+    },
+    {
+      "epoch": 129.9346405228758,
+      "grad_norm": 0.40253371000289917,
+      "learning_rate": 0.0001680922716037117,
+      "loss": 0.1219,
+      "step": 4970
+    },
+    {
+      "epoch": 130.19607843137254,
+      "grad_norm": 0.33488982915878296,
+      "learning_rate": 0.00016797108584307732,
+      "loss": 0.1202,
+      "step": 4980
+    },
+    {
+      "epoch": 130.45751633986927,
+      "grad_norm": 0.4206305146217346,
+      "learning_rate": 0.00016784971425208965,
+      "loss": 0.1189,
+      "step": 4990
+    },
+    {
+      "epoch": 130.718954248366,
+      "grad_norm": 0.4833426773548126,
+      "learning_rate": 0.00016772815716257412,
+      "loss": 0.1209,
+      "step": 5000
+    },
+    {
+      "epoch": 130.98039215686273,
+      "grad_norm": 0.5023657083511353,
+      "learning_rate": 0.00016760641490686307,
+      "loss": 0.1221,
+      "step": 5010
+    },
+    {
+      "epoch": 131.24183006535947,
+      "grad_norm": 0.4237312972545624,
+      "learning_rate": 0.0001674844878177952,
+      "loss": 0.1204,
+      "step": 5020
+    },
+    {
+      "epoch": 131.5032679738562,
+      "grad_norm": 0.42303699254989624,
+      "learning_rate": 0.00016736237622871452,
+      "loss": 0.1143,
+      "step": 5030
+    },
+    {
+      "epoch": 131.76470588235293,
+      "grad_norm": 0.4471609890460968,
+      "learning_rate": 0.00016724008047346947,
+      "loss": 0.1242,
+      "step": 5040
+    },
+    {
+      "epoch": 132.0261437908497,
+      "grad_norm": 0.43694472312927246,
+      "learning_rate": 0.00016711760088641196,
+      "loss": 0.1245,
+      "step": 5050
+    },
+    {
+      "epoch": 132.28758169934642,
+      "grad_norm": 0.460407018661499,
+      "learning_rate": 0.0001669949378023965,
+      "loss": 0.119,
+      "step": 5060
+    },
+    {
+      "epoch": 132.54901960784315,
+      "grad_norm": 0.39372462034225464,
+      "learning_rate": 0.00016687209155677929,
+      "loss": 0.1159,
+      "step": 5070
+    },
+    {
+      "epoch": 132.81045751633988,
+      "grad_norm": 0.38842689990997314,
+      "learning_rate": 0.00016674906248541726,
+      "loss": 0.1241,
+      "step": 5080
+    },
+    {
+      "epoch": 133.0718954248366,
+      "grad_norm": 0.40953928232192993,
+      "learning_rate": 0.00016662585092466723,
+      "loss": 0.1254,
+      "step": 5090
+    },
+    {
+      "epoch": 133.33333333333334,
+      "grad_norm": 0.4259697496891022,
+      "learning_rate": 0.0001665024572113848,
+      "loss": 0.1187,
+      "step": 5100
+    },
+    {
+      "epoch": 133.59477124183007,
+      "grad_norm": 0.46274760365486145,
+      "learning_rate": 0.00016637888168292384,
+      "loss": 0.1174,
+      "step": 5110
+    },
+    {
+      "epoch": 133.8562091503268,
+      "grad_norm": 0.5445423126220703,
+      "learning_rate": 0.000166255124677135,
+      "loss": 0.121,
+      "step": 5120
+    },
+    {
+      "epoch": 134.11764705882354,
+      "grad_norm": 0.38918498158454895,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 0.122,
+      "step": 5130
+    },
+    {
+      "epoch": 134.37908496732027,
+      "grad_norm": 0.3996954560279846,
+      "learning_rate": 0.00016600706758745668,
+      "loss": 0.1202,
+      "step": 5140
+    },
+    {
+      "epoch": 134.640522875817,
+      "grad_norm": 0.42393553256988525,
+      "learning_rate": 0.0001658827681817458,
+      "loss": 0.1226,
+      "step": 5150
+    },
+    {
+      "epoch": 134.90196078431373,
+      "grad_norm": 0.5191143751144409,
+      "learning_rate": 0.00016575828865506245,
+      "loss": 0.1208,
+      "step": 5160
+    },
+    {
+      "epoch": 135.16339869281046,
+      "grad_norm": 0.3241782784461975,
+      "learning_rate": 0.00016563362934772892,
+      "loss": 0.119,
+      "step": 5170
+    },
+    {
+      "epoch": 135.4248366013072,
+      "grad_norm": 0.4285520017147064,
+      "learning_rate": 0.00016550879060055895,
+      "loss": 0.1167,
+      "step": 5180
+    },
+    {
+      "epoch": 135.68627450980392,
+      "grad_norm": 0.4682152271270752,
+      "learning_rate": 0.00016538377275485691,
+      "loss": 0.1234,
+      "step": 5190
+    },
+    {
+      "epoch": 135.94771241830065,
+      "grad_norm": 0.5249207615852356,
+      "learning_rate": 0.00016525857615241687,
+      "loss": 0.122,
+      "step": 5200
+    },
+    {
+      "epoch": 136.20915032679738,
+      "grad_norm": 0.3867131173610687,
+      "learning_rate": 0.00016513320113552152,
+      "loss": 0.1183,
+      "step": 5210
+    },
+    {
+      "epoch": 136.47058823529412,
+      "grad_norm": 0.437995582818985,
+      "learning_rate": 0.0001650076480469413,
+      "loss": 0.1216,
+      "step": 5220
+    },
+    {
+      "epoch": 136.73202614379085,
+      "grad_norm": 0.47116023302078247,
+      "learning_rate": 0.0001648819172299337,
+      "loss": 0.1208,
+      "step": 5230
+    },
+    {
+      "epoch": 136.99346405228758,
+      "grad_norm": 0.4164656698703766,
+      "learning_rate": 0.0001647560090282419,
+      "loss": 0.1206,
+      "step": 5240
+    },
+    {
+      "epoch": 137.2549019607843,
+      "grad_norm": 0.41538143157958984,
+      "learning_rate": 0.00016462992378609407,
+      "loss": 0.1144,
+      "step": 5250
+    },
+    {
+      "epoch": 137.51633986928104,
+      "grad_norm": 0.48748481273651123,
+      "learning_rate": 0.00016450366184820255,
+      "loss": 0.1241,
+      "step": 5260
+    },
+    {
+      "epoch": 137.77777777777777,
+      "grad_norm": 0.49401819705963135,
+      "learning_rate": 0.00016437722355976258,
+      "loss": 0.1237,
+      "step": 5270
+    },
+    {
+      "epoch": 138.0392156862745,
+      "grad_norm": 0.41417965292930603,
+      "learning_rate": 0.00016425060926645167,
+      "loss": 0.1196,
+      "step": 5280
+    },
+    {
+      "epoch": 138.30065359477123,
+      "grad_norm": 0.34107911586761475,
+      "learning_rate": 0.00016412381931442838,
+      "loss": 0.1149,
+      "step": 5290
+    },
+    {
+      "epoch": 138.56209150326796,
+      "grad_norm": 0.5301646590232849,
+      "learning_rate": 0.00016399685405033167,
+      "loss": 0.1173,
+      "step": 5300
+    },
+    {
+      "epoch": 138.8235294117647,
+      "grad_norm": 0.500399649143219,
+      "learning_rate": 0.0001638697138212797,
+      "loss": 0.123,
+      "step": 5310
+    },
+    {
+      "epoch": 139.08496732026143,
+      "grad_norm": 0.43028515577316284,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.1254,
+      "step": 5320
+    },
+    {
+      "epoch": 139.34640522875816,
+      "grad_norm": 0.3889625668525696,
+      "learning_rate": 0.0001636149098591735,
+      "loss": 0.1163,
+      "step": 5330
+    },
+    {
+      "epoch": 139.6078431372549,
+      "grad_norm": 0.4060676693916321,
+      "learning_rate": 0.00016348724682274353,
+      "loss": 0.1185,
+      "step": 5340
+    },
+    {
+      "epoch": 139.86928104575162,
+      "grad_norm": 0.44942519068717957,
+      "learning_rate": 0.00016335941021460506,
+      "loss": 0.1216,
+      "step": 5350
+    },
+    {
+      "epoch": 140.13071895424838,
+      "grad_norm": 0.4668162167072296,
+      "learning_rate": 0.00016323140038425842,
+      "loss": 0.1222,
+      "step": 5360
+    },
+    {
+      "epoch": 140.3921568627451,
+      "grad_norm": 0.4359992444515228,
+      "learning_rate": 0.00016310321768167762,
+      "loss": 0.1183,
+      "step": 5370
+    },
+    {
+      "epoch": 140.65359477124184,
+      "grad_norm": 0.48698097467422485,
+      "learning_rate": 0.00016297486245730927,
+      "loss": 0.1182,
+      "step": 5380
+    },
+    {
+      "epoch": 140.91503267973857,
+      "grad_norm": 0.4088447391986847,
+      "learning_rate": 0.0001628463350620716,
+      "loss": 0.1259,
+      "step": 5390
+    },
+    {
+      "epoch": 141.1764705882353,
+      "grad_norm": 0.407049298286438,
+      "learning_rate": 0.0001627176358473537,
+      "loss": 0.1209,
+      "step": 5400
+    },
+    {
+      "epoch": 141.43790849673204,
+      "grad_norm": 0.4178103804588318,
+      "learning_rate": 0.00016258876516501424,
+      "loss": 0.1221,
+      "step": 5410
+    },
+    {
+      "epoch": 141.69934640522877,
+      "grad_norm": 0.44751349091529846,
+      "learning_rate": 0.0001624597233673808,
+      "loss": 0.1201,
+      "step": 5420
+    },
+    {
+      "epoch": 141.9607843137255,
+      "grad_norm": 0.4328495264053345,
+      "learning_rate": 0.00016233051080724868,
+      "loss": 0.1221,
+      "step": 5430
+    },
+    {
+      "epoch": 142.22222222222223,
+      "grad_norm": 0.44036683440208435,
+      "learning_rate": 0.0001622011278378801,
+      "loss": 0.1167,
+      "step": 5440
+    },
+    {
+      "epoch": 142.48366013071896,
+      "grad_norm": 0.5208015441894531,
+      "learning_rate": 0.00016207157481300312,
+      "loss": 0.1191,
+      "step": 5450
+    },
+    {
+      "epoch": 142.7450980392157,
+      "grad_norm": 0.4127715826034546,
+      "learning_rate": 0.00016194185208681083,
+      "loss": 0.1215,
+      "step": 5460
+    },
+    {
+      "epoch": 143.00653594771242,
+      "grad_norm": 0.4773133397102356,
+      "learning_rate": 0.00016181196001396019,
+      "loss": 0.1204,
+      "step": 5470
+    },
+    {
+      "epoch": 143.26797385620915,
+      "grad_norm": 0.37050458788871765,
+      "learning_rate": 0.0001616818989495711,
+      "loss": 0.1149,
+      "step": 5480
+    },
+    {
+      "epoch": 143.52941176470588,
+      "grad_norm": 0.34875085949897766,
+      "learning_rate": 0.00016155166924922566,
+      "loss": 0.1201,
+      "step": 5490
+    },
+    {
+      "epoch": 143.79084967320262,
+      "grad_norm": 0.3563850224018097,
+      "learning_rate": 0.0001614212712689668,
+      "loss": 0.1218,
+      "step": 5500
+    },
+    {
+      "epoch": 144.05228758169935,
+      "grad_norm": 0.37683019042015076,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 0.1199,
+      "step": 5510
+    },
+    {
+      "epoch": 144.31372549019608,
+      "grad_norm": 0.415547639131546,
+      "learning_rate": 0.00016115997189518043,
+      "loss": 0.1165,
+      "step": 5520
+    },
+    {
+      "epoch": 144.5751633986928,
+      "grad_norm": 0.3484375476837158,
+      "learning_rate": 0.00016102907121603543,
+      "loss": 0.1197,
+      "step": 5530
+    },
+    {
+      "epoch": 144.83660130718954,
+      "grad_norm": 0.3772566616535187,
+      "learning_rate": 0.00016089800368574014,
+      "loss": 0.1236,
+      "step": 5540
+    },
+    {
+      "epoch": 145.09803921568627,
+      "grad_norm": 0.35378995537757874,
+      "learning_rate": 0.00016076676966262813,
+      "loss": 0.1188,
+      "step": 5550
+    },
+    {
+      "epoch": 145.359477124183,
+      "grad_norm": 0.34689784049987793,
+      "learning_rate": 0.00016063536950548826,
+      "loss": 0.1169,
+      "step": 5560
+    },
+    {
+      "epoch": 145.62091503267973,
+      "grad_norm": 0.35840165615081787,
+      "learning_rate": 0.0001605038035735635,
+      "loss": 0.1199,
+      "step": 5570
+    },
+    {
+      "epoch": 145.88235294117646,
+      "grad_norm": 0.40955159068107605,
+      "learning_rate": 0.0001603720722265501,
+      "loss": 0.1236,
+      "step": 5580
+    },
+    {
+      "epoch": 146.1437908496732,
+      "grad_norm": 0.3746141493320465,
+      "learning_rate": 0.00016024017582459652,
+      "loss": 0.1176,
+      "step": 5590
+    },
+    {
+      "epoch": 146.40522875816993,
+      "grad_norm": 0.4733276069164276,
+      "learning_rate": 0.00016010811472830252,
+      "loss": 0.1198,
+      "step": 5600
+    },
+    {
+      "epoch": 146.66666666666666,
+      "grad_norm": 0.4011492431163788,
+      "learning_rate": 0.00015997588929871808,
+      "loss": 0.1199,
+      "step": 5610
+    },
+    {
+      "epoch": 146.9281045751634,
+      "grad_norm": 0.5024054050445557,
+      "learning_rate": 0.00015985674620589864,
+      "loss": 0.121,
+      "step": 5620
+    },
+    {
+      "epoch": 147.18954248366012,
+      "grad_norm": 0.4476403594017029,
+      "learning_rate": 0.00015972420953936335,
+      "loss": 0.1189,
+      "step": 5630
+    },
+    {
+      "epoch": 147.45098039215685,
+      "grad_norm": 0.45632895827293396,
+      "learning_rate": 0.0001595915095891198,
+      "loss": 0.1168,
+      "step": 5640
+    },
+    {
+      "epoch": 147.71241830065358,
+      "grad_norm": 0.48059502243995667,
+      "learning_rate": 0.00015945864671796452,
+      "loss": 0.1226,
+      "step": 5650
+    },
+    {
+      "epoch": 147.9738562091503,
+      "grad_norm": 0.3604694604873657,
+      "learning_rate": 0.0001593256212891395,
+      "loss": 0.1218,
+      "step": 5660
+    },
+    {
+      "epoch": 148.23529411764707,
+      "grad_norm": 1.3692547082901,
+      "learning_rate": 0.00015919243366633126,
+      "loss": 0.1168,
+      "step": 5670
+    },
+    {
+      "epoch": 148.4967320261438,
+      "grad_norm": 0.70684415102005,
+      "learning_rate": 0.00015905908421366962,
+      "loss": 0.1165,
+      "step": 5680
+    },
+    {
+      "epoch": 148.75816993464053,
+      "grad_norm": 0.9317061901092529,
+      "learning_rate": 0.0001589255732957269,
+      "loss": 0.1231,
+      "step": 5690
+    },
+    {
+      "epoch": 149.01960784313727,
+      "grad_norm": 0.6531818509101868,
+      "learning_rate": 0.00015879190127751684,
+      "loss": 0.1257,
+      "step": 5700
+    },
+    {
+      "epoch": 149.281045751634,
+      "grad_norm": 0.5326528549194336,
+      "learning_rate": 0.00015865806852449367,
+      "loss": 0.1136,
+      "step": 5710
+    },
+    {
+      "epoch": 149.54248366013073,
+      "grad_norm": 0.5003425478935242,
+      "learning_rate": 0.00015852407540255104,
+      "loss": 0.123,
+      "step": 5720
+    },
+    {
+      "epoch": 149.80392156862746,
+      "grad_norm": 0.4232426583766937,
+      "learning_rate": 0.00015838992227802093,
+      "loss": 0.1245,
+      "step": 5730
+    },
+    {
+      "epoch": 150.0653594771242,
+      "grad_norm": 0.3918308615684509,
+      "learning_rate": 0.00015825560951767298,
+      "loss": 0.118,
+      "step": 5740
+    },
+    {
+      "epoch": 150.32679738562092,
+      "grad_norm": 0.4566536545753479,
+      "learning_rate": 0.00015812113748871304,
+      "loss": 0.1194,
+      "step": 5750
+    },
+    {
+      "epoch": 150.58823529411765,
+      "grad_norm": 0.4792901277542114,
+      "learning_rate": 0.00015798650655878262,
+      "loss": 0.1235,
+      "step": 5760
+    },
+    {
+      "epoch": 150.84967320261438,
+      "grad_norm": 0.4351622462272644,
+      "learning_rate": 0.00015785171709595743,
+      "loss": 0.1183,
+      "step": 5770
+    },
+    {
+      "epoch": 151.11111111111111,
+      "grad_norm": 0.4398731589317322,
+      "learning_rate": 0.0001577167694687468,
+      "loss": 0.1213,
+      "step": 5780
+    },
+    {
+      "epoch": 151.37254901960785,
+      "grad_norm": 0.4514036476612091,
+      "learning_rate": 0.00015758166404609232,
+      "loss": 0.1184,
+      "step": 5790
+    },
+    {
+      "epoch": 151.63398692810458,
+      "grad_norm": 0.46276119351387024,
+      "learning_rate": 0.0001574464011973671,
+      "loss": 0.1182,
+      "step": 5800
+    },
+    {
+      "epoch": 151.8954248366013,
+      "grad_norm": 0.4101347327232361,
+      "learning_rate": 0.00015731098129237458,
+      "loss": 0.1245,
+      "step": 5810
+    },
+    {
+      "epoch": 152.15686274509804,
+      "grad_norm": 0.45118212699890137,
+      "learning_rate": 0.00015717540470134761,
+      "loss": 0.1173,
+      "step": 5820
+    },
+    {
+      "epoch": 152.41830065359477,
+      "grad_norm": 0.47819066047668457,
+      "learning_rate": 0.00015703967179494748,
+      "loss": 0.1193,
+      "step": 5830
+    },
+    {
+      "epoch": 152.6797385620915,
+      "grad_norm": 0.4761189818382263,
+      "learning_rate": 0.00015690378294426266,
+      "loss": 0.1208,
+      "step": 5840
+    },
+    {
+      "epoch": 152.94117647058823,
+      "grad_norm": 0.47973722219467163,
+      "learning_rate": 0.00015676773852080813,
+      "loss": 0.1205,
+      "step": 5850
+    },
+    {
+      "epoch": 153.20261437908496,
+      "grad_norm": 0.4598498046398163,
+      "learning_rate": 0.0001566315388965242,
+      "loss": 0.119,
+      "step": 5860
+    },
+    {
+      "epoch": 153.4640522875817,
+      "grad_norm": 0.42741018533706665,
+      "learning_rate": 0.00015649518444377537,
+      "loss": 0.1176,
+      "step": 5870
+    },
+    {
+      "epoch": 153.72549019607843,
+      "grad_norm": 0.40698984265327454,
+      "learning_rate": 0.00015635867553534955,
+      "loss": 0.1215,
+      "step": 5880
+    },
+    {
+      "epoch": 153.98692810457516,
+      "grad_norm": 0.35974419116973877,
+      "learning_rate": 0.00015622201254445684,
+      "loss": 0.1221,
+      "step": 5890
+    },
+    {
+      "epoch": 154.2483660130719,
+      "grad_norm": 0.4019433856010437,
+      "learning_rate": 0.0001560851958447287,
+      "loss": 0.1168,
+      "step": 5900
+    },
+    {
+      "epoch": 154.50980392156862,
+      "grad_norm": 0.4910339117050171,
+      "learning_rate": 0.0001559482258102167,
+      "loss": 0.1213,
+      "step": 5910
+    },
+    {
+      "epoch": 154.77124183006535,
+      "grad_norm": 0.40200385451316833,
+      "learning_rate": 0.00015581110281539173,
+      "loss": 0.12,
+      "step": 5920
+    },
+    {
+      "epoch": 155.03267973856208,
+      "grad_norm": 0.36135050654411316,
+      "learning_rate": 0.0001556738272351428,
+      "loss": 0.1166,
+      "step": 5930
+    },
+    {
+      "epoch": 155.2941176470588,
+      "grad_norm": 0.41371849179267883,
+      "learning_rate": 0.00015553639944477612,
+      "loss": 0.1184,
+      "step": 5940
+    },
+    {
+      "epoch": 155.55555555555554,
+      "grad_norm": 0.4281088411808014,
+      "learning_rate": 0.000155398819820014,
+      "loss": 0.1187,
+      "step": 5950
+    },
+    {
+      "epoch": 155.81699346405227,
+      "grad_norm": 0.530092716217041,
+      "learning_rate": 0.00015526108873699387,
+      "loss": 0.1203,
+      "step": 5960
+    },
+    {
+      "epoch": 156.07843137254903,
+      "grad_norm": 0.3755616843700409,
+      "learning_rate": 0.00015512320657226728,
+      "loss": 0.1175,
+      "step": 5970
+    },
+    {
+      "epoch": 156.33986928104576,
+      "grad_norm": 0.35955289006233215,
+      "learning_rate": 0.00015498517370279884,
+      "loss": 0.1173,
+      "step": 5980
+    },
+    {
+      "epoch": 156.6013071895425,
+      "grad_norm": 0.406409353017807,
+      "learning_rate": 0.00015484699050596505,
+      "loss": 0.1221,
+      "step": 5990
+    },
+    {
+      "epoch": 156.86274509803923,
+      "grad_norm": 0.4339780807495117,
+      "learning_rate": 0.00015470865735955357,
+      "loss": 0.1183,
+      "step": 6000
+    },
+    {
+      "epoch": 157.12418300653596,
+      "grad_norm": 0.35623541474342346,
+      "learning_rate": 0.00015457017464176191,
+      "loss": 0.1215,
+      "step": 6010
+    },
+    {
+      "epoch": 157.3856209150327,
+      "grad_norm": 0.43511274456977844,
+      "learning_rate": 0.0001544315427311965,
+      "loss": 0.1169,
+      "step": 6020
+    },
+    {
+      "epoch": 157.64705882352942,
+      "grad_norm": 0.4758042097091675,
+      "learning_rate": 0.00015429276200687177,
+      "loss": 0.1177,
+      "step": 6030
+    },
+    {
+      "epoch": 157.90849673202615,
+      "grad_norm": 0.3972184360027313,
+      "learning_rate": 0.00015415383284820888,
+      "loss": 0.123,
+      "step": 6040
+    },
+    {
+      "epoch": 158.16993464052288,
+      "grad_norm": 0.3831634223461151,
+      "learning_rate": 0.0001540147556350348,
+      "loss": 0.1187,
+      "step": 6050
+    },
+    {
+      "epoch": 158.4313725490196,
+      "grad_norm": 0.3707033097743988,
+      "learning_rate": 0.0001538755307475814,
+      "loss": 0.1166,
+      "step": 6060
+    },
+    {
+      "epoch": 158.69281045751634,
+      "grad_norm": 0.40588223934173584,
+      "learning_rate": 0.00015373615856648418,
+      "loss": 0.1201,
+      "step": 6070
+    },
+    {
+      "epoch": 158.95424836601308,
+      "grad_norm": 0.5210367441177368,
+      "learning_rate": 0.0001535966394727813,
+      "loss": 0.1223,
+      "step": 6080
+    },
+    {
+      "epoch": 159.2156862745098,
+      "grad_norm": 0.3465724587440491,
+      "learning_rate": 0.00015345697384791274,
+      "loss": 0.1162,
+      "step": 6090
+    },
+    {
+      "epoch": 159.47712418300654,
+      "grad_norm": 0.48409783840179443,
+      "learning_rate": 0.00015331716207371888,
+      "loss": 0.1186,
+      "step": 6100
+    },
+    {
+      "epoch": 159.73856209150327,
+      "grad_norm": 0.49288466572761536,
+      "learning_rate": 0.00015317720453243981,
+      "loss": 0.1201,
+      "step": 6110
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 0.41373661160469055,
+      "learning_rate": 0.00015303710160671416,
+      "loss": 0.1221,
+      "step": 6120
+    },
+    {
+      "epoch": 160.26143790849673,
+      "grad_norm": 0.30801451206207275,
+      "learning_rate": 0.00015289685367957792,
+      "loss": 0.1187,
+      "step": 6130
+    },
+    {
+      "epoch": 160.52287581699346,
+      "grad_norm": 0.4610970616340637,
+      "learning_rate": 0.0001527564611344636,
+      "loss": 0.1183,
+      "step": 6140
+    },
+    {
+      "epoch": 160.7843137254902,
+      "grad_norm": 0.3737730383872986,
+      "learning_rate": 0.000152615924355199,
+      "loss": 0.1174,
+      "step": 6150
+    },
+    {
+      "epoch": 161.04575163398692,
+      "grad_norm": 0.4367047846317291,
+      "learning_rate": 0.00015247524372600637,
+      "loss": 0.1245,
+      "step": 6160
+    },
+    {
+      "epoch": 161.30718954248366,
+      "grad_norm": 0.4288366734981537,
+      "learning_rate": 0.00015233441963150113,
+      "loss": 0.1165,
+      "step": 6170
+    },
+    {
+      "epoch": 161.5686274509804,
+      "grad_norm": 0.5186516642570496,
+      "learning_rate": 0.00015219345245669105,
+      "loss": 0.1196,
+      "step": 6180
+    },
+    {
+      "epoch": 161.83006535947712,
+      "grad_norm": 0.32446616888046265,
+      "learning_rate": 0.00015205234258697494,
+      "loss": 0.1206,
+      "step": 6190
+    },
+    {
+      "epoch": 162.09150326797385,
+      "grad_norm": 0.34923261404037476,
+      "learning_rate": 0.00015191109040814176,
+      "loss": 0.1184,
+      "step": 6200
+    },
+    {
+      "epoch": 162.35294117647058,
+      "grad_norm": 0.4656030833721161,
+      "learning_rate": 0.0001517696963063697,
+      "loss": 0.1186,
+      "step": 6210
+    },
+    {
+      "epoch": 162.6143790849673,
+      "grad_norm": 0.3549903631210327,
+      "learning_rate": 0.0001516281606682247,
+      "loss": 0.1204,
+      "step": 6220
+    },
+    {
+      "epoch": 162.87581699346404,
+      "grad_norm": 0.3952346742153168,
+      "learning_rate": 0.0001514864838806599,
+      "loss": 0.1161,
+      "step": 6230
+    },
+    {
+      "epoch": 163.13725490196077,
+      "grad_norm": 0.3911607563495636,
+      "learning_rate": 0.0001513446663310141,
+      "loss": 0.1198,
+      "step": 6240
+    },
+    {
+      "epoch": 163.3986928104575,
+      "grad_norm": 0.346305787563324,
+      "learning_rate": 0.00015120270840701124,
+      "loss": 0.1118,
+      "step": 6250
+    },
+    {
+      "epoch": 163.66013071895424,
+      "grad_norm": 0.42828211188316345,
+      "learning_rate": 0.0001510606104967587,
+      "loss": 0.1195,
+      "step": 6260
+    },
+    {
+      "epoch": 163.92156862745097,
+      "grad_norm": 0.4532879590988159,
+      "learning_rate": 0.00015091837298874682,
+      "loss": 0.1241,
+      "step": 6270
+    },
+    {
+      "epoch": 164.18300653594773,
+      "grad_norm": 0.33145442605018616,
+      "learning_rate": 0.00015077599627184754,
+      "loss": 0.116,
+      "step": 6280
+    },
+    {
+      "epoch": 164.44444444444446,
+      "grad_norm": 0.4350489675998688,
+      "learning_rate": 0.00015063348073531324,
+      "loss": 0.1173,
+      "step": 6290
+    },
+    {
+      "epoch": 164.7058823529412,
+      "grad_norm": 0.4127410054206848,
+      "learning_rate": 0.00015049082676877614,
+      "loss": 0.1209,
+      "step": 6300
+    },
+    {
+      "epoch": 164.96732026143792,
+      "grad_norm": 0.36359113454818726,
+      "learning_rate": 0.00015034803476224657,
+      "loss": 0.1215,
+      "step": 6310
+    },
+    {
+      "epoch": 165.22875816993465,
+      "grad_norm": 0.3537936210632324,
+      "learning_rate": 0.00015020510510611255,
+      "loss": 0.1203,
+      "step": 6320
+    },
+    {
+      "epoch": 165.49019607843138,
+      "grad_norm": 0.33729109168052673,
+      "learning_rate": 0.00015006203819113823,
+      "loss": 0.1162,
+      "step": 6330
+    },
+    {
+      "epoch": 165.7516339869281,
+      "grad_norm": 0.4218641221523285,
+      "learning_rate": 0.00014991883440846308,
+      "loss": 0.1162,
+      "step": 6340
+    },
+    {
+      "epoch": 166.01307189542484,
+      "grad_norm": 0.44932490587234497,
+      "learning_rate": 0.00014977549414960084,
+      "loss": 0.1243,
+      "step": 6350
+    },
+    {
+      "epoch": 166.27450980392157,
+      "grad_norm": 0.430530846118927,
+      "learning_rate": 0.00014963201780643823,
+      "loss": 0.1167,
+      "step": 6360
+    },
+    {
+      "epoch": 166.5359477124183,
+      "grad_norm": 0.4142252504825592,
+      "learning_rate": 0.00014948840577123416,
+      "loss": 0.1172,
+      "step": 6370
+    },
+    {
+      "epoch": 166.79738562091504,
+      "grad_norm": 0.2989576458930969,
+      "learning_rate": 0.00014934465843661842,
+      "loss": 0.1197,
+      "step": 6380
+    },
+    {
+      "epoch": 167.05882352941177,
+      "grad_norm": 0.37099623680114746,
+      "learning_rate": 0.00014920077619559073,
+      "loss": 0.1207,
+      "step": 6390
+    },
+    {
+      "epoch": 167.3202614379085,
+      "grad_norm": 0.3625079393386841,
+      "learning_rate": 0.00014905675944151966,
+      "loss": 0.1186,
+      "step": 6400
+    },
+    {
+      "epoch": 167.58169934640523,
+      "grad_norm": 0.4216829836368561,
+      "learning_rate": 0.00014891260856814148,
+      "loss": 0.1154,
+      "step": 6410
+    },
+    {
+      "epoch": 167.84313725490196,
+      "grad_norm": 0.44304612278938293,
+      "learning_rate": 0.0001487683239695592,
+      "loss": 0.1211,
+      "step": 6420
+    },
+    {
+      "epoch": 168.1045751633987,
+      "grad_norm": 0.38862261176109314,
+      "learning_rate": 0.00014862390604024144,
+      "loss": 0.1166,
+      "step": 6430
+    },
+    {
+      "epoch": 168.36601307189542,
+      "grad_norm": 0.38382863998413086,
+      "learning_rate": 0.00014847935517502123,
+      "loss": 0.1139,
+      "step": 6440
+    },
+    {
+      "epoch": 168.62745098039215,
+      "grad_norm": 0.42605075240135193,
+      "learning_rate": 0.00014833467176909515,
+      "loss": 0.1209,
+      "step": 6450
+    },
+    {
+      "epoch": 168.88888888888889,
+      "grad_norm": 0.3948115408420563,
+      "learning_rate": 0.00014818985621802212,
+      "loss": 0.12,
+      "step": 6460
+    },
+    {
+      "epoch": 169.15032679738562,
+      "grad_norm": 0.33048856258392334,
+      "learning_rate": 0.00014804490891772232,
+      "loss": 0.1185,
+      "step": 6470
+    },
+    {
+      "epoch": 169.41176470588235,
+      "grad_norm": 0.33838731050491333,
+      "learning_rate": 0.00014789983026447612,
+      "loss": 0.1139,
+      "step": 6480
+    },
+    {
+      "epoch": 169.67320261437908,
+      "grad_norm": 0.40736544132232666,
+      "learning_rate": 0.000147754620654923,
+      "loss": 0.1212,
+      "step": 6490
+    },
+    {
+      "epoch": 169.9346405228758,
+      "grad_norm": 0.5088180899620056,
+      "learning_rate": 0.00014760928048606055,
+      "loss": 0.1211,
+      "step": 6500
+    },
+    {
+      "epoch": 170.19607843137254,
+      "grad_norm": 0.36517319083213806,
+      "learning_rate": 0.00014746381015524323,
+      "loss": 0.1204,
+      "step": 6510
+    },
+    {
+      "epoch": 170.45751633986927,
+      "grad_norm": 0.4173334836959839,
+      "learning_rate": 0.00014731821006018131,
+      "loss": 0.1138,
+      "step": 6520
+    },
+    {
+      "epoch": 170.718954248366,
+      "grad_norm": 0.3489775061607361,
+      "learning_rate": 0.00014717248059893992,
+      "loss": 0.1197,
+      "step": 6530
+    },
+    {
+      "epoch": 170.98039215686273,
+      "grad_norm": 0.3684404492378235,
+      "learning_rate": 0.00014702662216993785,
+      "loss": 0.1226,
+      "step": 6540
+    },
+    {
+      "epoch": 171.24183006535947,
+      "grad_norm": 0.3432365953922272,
+      "learning_rate": 0.0001468806351719465,
+      "loss": 0.1186,
+      "step": 6550
+    },
+    {
+      "epoch": 171.5032679738562,
+      "grad_norm": 0.4222089350223541,
+      "learning_rate": 0.0001467345200040887,
+      "loss": 0.1193,
+      "step": 6560
+    },
+    {
+      "epoch": 171.76470588235293,
+      "grad_norm": 0.43939515948295593,
+      "learning_rate": 0.0001465882770658378,
+      "loss": 0.1197,
+      "step": 6570
+    },
+    {
+      "epoch": 172.0261437908497,
+      "grad_norm": 0.3959405720233917,
+      "learning_rate": 0.00014644190675701632,
+      "loss": 0.1188,
+      "step": 6580
+    },
+    {
+      "epoch": 172.28758169934642,
+      "grad_norm": 0.3560381531715393,
+      "learning_rate": 0.00014629540947779516,
+      "loss": 0.111,
+      "step": 6590
+    },
+    {
+      "epoch": 172.54901960784315,
+      "grad_norm": 0.31908681988716125,
+      "learning_rate": 0.0001461487856286923,
+      "loss": 0.1179,
+      "step": 6600
+    },
+    {
+      "epoch": 172.81045751633988,
+      "grad_norm": 0.4550021290779114,
+      "learning_rate": 0.0001460020356105717,
+      "loss": 0.1229,
+      "step": 6610
+    },
+    {
+      "epoch": 173.0718954248366,
+      "grad_norm": 0.31013500690460205,
+      "learning_rate": 0.00014585515982464234,
+      "loss": 0.1192,
+      "step": 6620
+    },
+    {
+      "epoch": 173.33333333333334,
+      "grad_norm": 0.41469523310661316,
+      "learning_rate": 0.00014570815867245696,
+      "loss": 0.1183,
+      "step": 6630
+    },
+    {
+      "epoch": 173.59477124183007,
+      "grad_norm": 0.40070420503616333,
+      "learning_rate": 0.00014556103255591114,
+      "loss": 0.1177,
+      "step": 6640
+    },
+    {
+      "epoch": 173.8562091503268,
+      "grad_norm": 0.3658972978591919,
+      "learning_rate": 0.0001454137818772421,
+      "loss": 0.1203,
+      "step": 6650
+    },
+    {
+      "epoch": 174.11764705882354,
+      "grad_norm": 0.4114379584789276,
+      "learning_rate": 0.00014526640703902747,
+      "loss": 0.1159,
+      "step": 6660
+    },
+    {
+      "epoch": 174.37908496732027,
+      "grad_norm": 0.37510064244270325,
+      "learning_rate": 0.00014511890844418453,
+      "loss": 0.1205,
+      "step": 6670
+    },
+    {
+      "epoch": 174.640522875817,
+      "grad_norm": 0.41856005787849426,
+      "learning_rate": 0.00014497128649596875,
+      "loss": 0.1169,
+      "step": 6680
+    },
+    {
+      "epoch": 174.90196078431373,
+      "grad_norm": 0.35670095682144165,
+      "learning_rate": 0.00014482354159797288,
+      "loss": 0.1187,
+      "step": 6690
+    },
+    {
+      "epoch": 175.16339869281046,
+      "grad_norm": 0.3349741995334625,
+      "learning_rate": 0.0001446756741541259,
+      "loss": 0.1191,
+      "step": 6700
+    },
+    {
+      "epoch": 175.4248366013072,
+      "grad_norm": 0.4219711422920227,
+      "learning_rate": 0.00014452768456869173,
+      "loss": 0.1167,
+      "step": 6710
+    },
+    {
+      "epoch": 175.68627450980392,
+      "grad_norm": 0.30478203296661377,
+      "learning_rate": 0.0001443795732462682,
+      "loss": 0.1172,
+      "step": 6720
+    },
+    {
+      "epoch": 175.94771241830065,
+      "grad_norm": 0.3981144428253174,
+      "learning_rate": 0.00014423134059178607,
+      "loss": 0.121,
+      "step": 6730
+    },
+    {
+      "epoch": 176.20915032679738,
+      "grad_norm": 0.3811021149158478,
+      "learning_rate": 0.00014408298701050774,
+      "loss": 0.1182,
+      "step": 6740
+    },
+    {
+      "epoch": 176.47058823529412,
+      "grad_norm": 0.4203794002532959,
+      "learning_rate": 0.00014393451290802619,
+      "loss": 0.1191,
+      "step": 6750
+    },
+    {
+      "epoch": 176.73202614379085,
+      "grad_norm": 0.38078659772872925,
+      "learning_rate": 0.000143785918690264,
+      "loss": 0.117,
+      "step": 6760
+    },
+    {
+      "epoch": 176.99346405228758,
+      "grad_norm": 0.3496187627315521,
+      "learning_rate": 0.0001436372047634721,
+      "loss": 0.1197,
+      "step": 6770
+    },
+    {
+      "epoch": 177.2549019607843,
+      "grad_norm": 0.3604697287082672,
+      "learning_rate": 0.00014348837153422864,
+      "loss": 0.1182,
+      "step": 6780
+    },
+    {
+      "epoch": 177.51633986928104,
+      "grad_norm": 0.3979366421699524,
+      "learning_rate": 0.000143339419409438,
+      "loss": 0.1169,
+      "step": 6790
+    },
+    {
+      "epoch": 177.77777777777777,
+      "grad_norm": 0.4077358543872833,
+      "learning_rate": 0.00014319034879632962,
+      "loss": 0.1189,
+      "step": 6800
+    },
+    {
+      "epoch": 178.0392156862745,
+      "grad_norm": 0.377392053604126,
+      "learning_rate": 0.00014304116010245685,
+      "loss": 0.1205,
+      "step": 6810
+    },
+    {
+      "epoch": 178.30065359477123,
+      "grad_norm": 0.9518114924430847,
+      "learning_rate": 0.00014289185373569585,
+      "loss": 0.1185,
+      "step": 6820
+    },
+    {
+      "epoch": 178.56209150326796,
+      "grad_norm": 0.4851488173007965,
+      "learning_rate": 0.00014274243010424457,
+      "loss": 0.1222,
+      "step": 6830
+    },
+    {
+      "epoch": 178.8235294117647,
+      "grad_norm": 0.4688913822174072,
+      "learning_rate": 0.00014259288961662153,
+      "loss": 0.1211,
+      "step": 6840
+    },
+    {
+      "epoch": 179.08496732026143,
+      "grad_norm": 0.4016769528388977,
+      "learning_rate": 0.00014244323268166467,
+      "loss": 0.1216,
+      "step": 6850
+    },
+    {
+      "epoch": 179.34640522875816,
+      "grad_norm": 0.661056637763977,
+      "learning_rate": 0.00014229345970853032,
+      "loss": 0.1203,
+      "step": 6860
+    },
+    {
+      "epoch": 179.6078431372549,
+      "grad_norm": 0.8823976516723633,
+      "learning_rate": 0.00014214357110669211,
+      "loss": 0.1218,
+      "step": 6870
+    },
+    {
+      "epoch": 179.86928104575162,
+      "grad_norm": 0.6820003390312195,
+      "learning_rate": 0.00014199356728593977,
+      "loss": 0.1225,
+      "step": 6880
+    },
+    {
+      "epoch": 180.13071895424838,
+      "grad_norm": 1.0853325128555298,
+      "learning_rate": 0.000141843448656378,
+      "loss": 0.1259,
+      "step": 6890
+    },
+    {
+      "epoch": 180.3921568627451,
+      "grad_norm": 0.9575023055076599,
+      "learning_rate": 0.00014169321562842535,
+      "loss": 0.1179,
+      "step": 6900
+    },
+    {
+      "epoch": 180.65359477124184,
+      "grad_norm": 0.4306863248348236,
+      "learning_rate": 0.00014154286861281325,
+      "loss": 0.124,
+      "step": 6910
+    },
+    {
+      "epoch": 180.91503267973857,
+      "grad_norm": 0.48212000727653503,
+      "learning_rate": 0.00014139240802058464,
+      "loss": 0.1216,
+      "step": 6920
+    },
+    {
+      "epoch": 181.1764705882353,
+      "grad_norm": 0.7190878391265869,
+      "learning_rate": 0.0001412418342630931,
+      "loss": 0.121,
+      "step": 6930
+    },
+    {
+      "epoch": 181.43790849673204,
+      "grad_norm": 0.6326178908348083,
+      "learning_rate": 0.0001410911477520015,
+      "loss": 0.122,
+      "step": 6940
+    },
+    {
+      "epoch": 181.69934640522877,
+      "grad_norm": 0.6375879645347595,
+      "learning_rate": 0.000140940348899281,
+      "loss": 0.1231,
+      "step": 6950
+    },
+    {
+      "epoch": 181.9607843137255,
+      "grad_norm": 0.8026982545852661,
+      "learning_rate": 0.0001407894381172099,
+      "loss": 0.1278,
+      "step": 6960
+    },
+    {
+      "epoch": 182.22222222222223,
+      "grad_norm": 0.8561803102493286,
+      "learning_rate": 0.00014063841581837255,
+      "loss": 0.1169,
+      "step": 6970
+    },
+    {
+      "epoch": 182.48366013071896,
+      "grad_norm": 0.6810368895530701,
+      "learning_rate": 0.00014048728241565812,
+      "loss": 0.1255,
+      "step": 6980
+    },
+    {
+      "epoch": 182.7450980392157,
+      "grad_norm": 0.8623539209365845,
+      "learning_rate": 0.00014033603832225956,
+      "loss": 0.1241,
+      "step": 6990
+    },
+    {
+      "epoch": 183.00653594771242,
+      "grad_norm": 0.5568687915802002,
+      "learning_rate": 0.00014018468395167246,
+      "loss": 0.1291,
+      "step": 7000
+    },
+    {
+      "epoch": 183.26797385620915,
+      "grad_norm": 0.7413952946662903,
+      "learning_rate": 0.00014003321971769385,
+      "loss": 0.1186,
+      "step": 7010
+    },
+    {
+      "epoch": 183.52941176470588,
+      "grad_norm": 0.39381900429725647,
+      "learning_rate": 0.00013988164603442126,
+      "loss": 0.1181,
+      "step": 7020
+    },
+    {
+      "epoch": 183.79084967320262,
+      "grad_norm": 0.8067420721054077,
+      "learning_rate": 0.00013972996331625126,
+      "loss": 0.1233,
+      "step": 7030
+    },
+    {
+      "epoch": 184.05228758169935,
+      "grad_norm": 0.4819212257862091,
+      "learning_rate": 0.00013957817197787865,
+      "loss": 0.1228,
+      "step": 7040
+    },
+    {
+      "epoch": 184.31372549019608,
+      "grad_norm": 0.5741551518440247,
+      "learning_rate": 0.00013942627243429512,
+      "loss": 0.1186,
+      "step": 7050
+    },
+    {
+      "epoch": 184.5751633986928,
+      "grad_norm": 0.44853687286376953,
+      "learning_rate": 0.0001392742651007882,
+      "loss": 0.1195,
+      "step": 7060
+    },
+    {
+      "epoch": 184.83660130718954,
+      "grad_norm": 0.39852526783943176,
+      "learning_rate": 0.00013912215039294028,
+      "loss": 0.1197,
+      "step": 7070
+    },
+    {
+      "epoch": 185.09803921568627,
+      "grad_norm": 0.3141544461250305,
+      "learning_rate": 0.000138969928726627,
+      "loss": 0.1196,
+      "step": 7080
+    },
+    {
+      "epoch": 185.359477124183,
+      "grad_norm": 0.40422919392585754,
+      "learning_rate": 0.00013881760051801667,
+      "loss": 0.117,
+      "step": 7090
+    },
+    {
+      "epoch": 185.62091503267973,
+      "grad_norm": 0.38428983092308044,
+      "learning_rate": 0.00013866516618356875,
+      "loss": 0.1165,
+      "step": 7100
+    },
+    {
+      "epoch": 185.88235294117646,
+      "grad_norm": 0.42969101667404175,
+      "learning_rate": 0.00013851262614003292,
+      "loss": 0.1211,
+      "step": 7110
+    },
+    {
+      "epoch": 186.1437908496732,
+      "grad_norm": 0.35507911443710327,
+      "learning_rate": 0.0001383599808044478,
+      "loss": 0.1167,
+      "step": 7120
+    },
+    {
+      "epoch": 186.40522875816993,
+      "grad_norm": 0.3685397207736969,
+      "learning_rate": 0.00013820723059413995,
+      "loss": 0.1169,
+      "step": 7130
+    },
+    {
+      "epoch": 186.66666666666666,
+      "grad_norm": 0.4494422972202301,
+      "learning_rate": 0.00013805437592672262,
+      "loss": 0.1144,
+      "step": 7140
+    },
+    {
+      "epoch": 186.9281045751634,
+      "grad_norm": 0.5370398759841919,
+      "learning_rate": 0.00013790141722009458,
+      "loss": 0.1222,
+      "step": 7150
+    },
+    {
+      "epoch": 187.18954248366012,
+      "grad_norm": 0.4284844994544983,
+      "learning_rate": 0.00013774835489243912,
+      "loss": 0.1185,
+      "step": 7160
+    },
+    {
+      "epoch": 187.45098039215685,
+      "grad_norm": 0.5446126461029053,
+      "learning_rate": 0.0001375951893622228,
+      "loss": 0.1202,
+      "step": 7170
+    },
+    {
+      "epoch": 187.71241830065358,
+      "grad_norm": 0.388539582490921,
+      "learning_rate": 0.00013744192104819437,
+      "loss": 0.118,
+      "step": 7180
+    },
+    {
+      "epoch": 187.9738562091503,
+      "grad_norm": 0.3365419805049896,
+      "learning_rate": 0.00013728855036938348,
+      "loss": 0.1169,
+      "step": 7190
+    },
+    {
+      "epoch": 188.23529411764707,
+      "grad_norm": 0.36737489700317383,
+      "learning_rate": 0.00013713507774509973,
+      "loss": 0.1145,
+      "step": 7200
+    },
+    {
+      "epoch": 188.4967320261438,
+      "grad_norm": 0.44414252042770386,
+      "learning_rate": 0.00013698150359493143,
+      "loss": 0.1166,
+      "step": 7210
+    },
+    {
+      "epoch": 188.75816993464053,
+      "grad_norm": 0.3925827443599701,
+      "learning_rate": 0.00013682782833874442,
+      "loss": 0.1198,
+      "step": 7220
+    },
+    {
+      "epoch": 189.01960784313727,
+      "grad_norm": 0.46063002943992615,
+      "learning_rate": 0.00013667405239668106,
+      "loss": 0.1211,
+      "step": 7230
+    },
+    {
+      "epoch": 189.281045751634,
+      "grad_norm": 0.33849823474884033,
+      "learning_rate": 0.0001365201761891588,
+      "loss": 0.1114,
+      "step": 7240
+    },
+    {
+      "epoch": 189.54248366013073,
+      "grad_norm": 0.4201454222202301,
+      "learning_rate": 0.00013636620013686936,
+      "loss": 0.1176,
+      "step": 7250
+    },
+    {
+      "epoch": 189.80392156862746,
+      "grad_norm": 0.3787872791290283,
+      "learning_rate": 0.00013621212466077736,
+      "loss": 0.1181,
+      "step": 7260
+    },
+    {
+      "epoch": 190.0653594771242,
+      "grad_norm": 0.3665624260902405,
+      "learning_rate": 0.00013605795018211932,
+      "loss": 0.1196,
+      "step": 7270
+    },
+    {
+      "epoch": 190.32679738562092,
+      "grad_norm": 0.3760620355606079,
+      "learning_rate": 0.0001359036771224024,
+      "loss": 0.1164,
+      "step": 7280
+    },
+    {
+      "epoch": 190.58823529411765,
+      "grad_norm": 0.4009071886539459,
+      "learning_rate": 0.00013574930590340314,
+      "loss": 0.1208,
+      "step": 7290
+    },
+    {
+      "epoch": 190.84967320261438,
+      "grad_norm": 0.3013065755367279,
+      "learning_rate": 0.00013559483694716663,
+      "loss": 0.119,
+      "step": 7300
+    },
+    {
+      "epoch": 191.11111111111111,
+      "grad_norm": 0.3610047996044159,
+      "learning_rate": 0.00013544027067600512,
+      "loss": 0.115,
+      "step": 7310
+    },
+    {
+      "epoch": 191.37254901960785,
+      "grad_norm": 0.3881337344646454,
+      "learning_rate": 0.00013528560751249687,
+      "loss": 0.1157,
+      "step": 7320
+    },
+    {
+      "epoch": 191.63398692810458,
+      "grad_norm": 0.36018821597099304,
+      "learning_rate": 0.00013513084787948504,
+      "loss": 0.1158,
+      "step": 7330
+    },
+    {
+      "epoch": 191.8954248366013,
+      "grad_norm": 0.4123559296131134,
+      "learning_rate": 0.00013497599220007656,
+      "loss": 0.1193,
+      "step": 7340
+    },
+    {
+      "epoch": 192.15686274509804,
+      "grad_norm": 0.35036417841911316,
+      "learning_rate": 0.00013482104089764096,
+      "loss": 0.117,
+      "step": 7350
+    },
+    {
+      "epoch": 192.41830065359477,
+      "grad_norm": 0.4409710168838501,
+      "learning_rate": 0.0001346659943958092,
+      "loss": 0.1147,
+      "step": 7360
+    },
+    {
+      "epoch": 192.6797385620915,
+      "grad_norm": 0.3965952694416046,
+      "learning_rate": 0.0001345108531184725,
+      "loss": 0.1178,
+      "step": 7370
+    },
+    {
+      "epoch": 192.94117647058823,
+      "grad_norm": 0.34038016200065613,
+      "learning_rate": 0.00013435561748978113,
+      "loss": 0.1216,
+      "step": 7380
+    },
+    {
+      "epoch": 193.20261437908496,
+      "grad_norm": 0.30114054679870605,
+      "learning_rate": 0.00013420028793414344,
+      "loss": 0.1153,
+      "step": 7390
+    },
+    {
+      "epoch": 193.4640522875817,
+      "grad_norm": 0.34901145100593567,
+      "learning_rate": 0.00013404486487622442,
+      "loss": 0.1176,
+      "step": 7400
+    },
+    {
+      "epoch": 193.72549019607843,
+      "grad_norm": 0.4058627784252167,
+      "learning_rate": 0.00013388934874094489,
+      "loss": 0.1168,
+      "step": 7410
+    },
+    {
+      "epoch": 193.98692810457516,
+      "grad_norm": 0.31916511058807373,
+      "learning_rate": 0.00013373373995347995,
+      "loss": 0.1214,
+      "step": 7420
+    },
+    {
+      "epoch": 194.2483660130719,
+      "grad_norm": 0.3582738935947418,
+      "learning_rate": 0.00013357803893925807,
+      "loss": 0.1155,
+      "step": 7430
+    },
+    {
+      "epoch": 194.50980392156862,
+      "grad_norm": 0.3602858781814575,
+      "learning_rate": 0.00013342224612395993,
+      "loss": 0.121,
+      "step": 7440
+    },
+    {
+      "epoch": 194.77124183006535,
+      "grad_norm": 0.30642154812812805,
+      "learning_rate": 0.0001332663619335171,
+      "loss": 0.1149,
+      "step": 7450
+    },
+    {
+      "epoch": 195.03267973856208,
+      "grad_norm": 0.32698705792427063,
+      "learning_rate": 0.00013311038679411104,
+      "loss": 0.1184,
+      "step": 7460
+    },
+    {
+      "epoch": 195.2941176470588,
+      "grad_norm": 0.34545981884002686,
+      "learning_rate": 0.00013295432113217176,
+      "loss": 0.1139,
+      "step": 7470
+    },
+    {
+      "epoch": 195.55555555555554,
+      "grad_norm": 0.2823973298072815,
+      "learning_rate": 0.00013279816537437687,
+      "loss": 0.1156,
+      "step": 7480
+    },
+    {
+      "epoch": 195.81699346405227,
+      "grad_norm": 0.3787044882774353,
+      "learning_rate": 0.00013264191994765028,
+      "loss": 0.1199,
+      "step": 7490
+    },
+    {
+      "epoch": 196.07843137254903,
+      "grad_norm": 0.2924850583076477,
+      "learning_rate": 0.00013248558527916094,
+      "loss": 0.1169,
+      "step": 7500
+    },
+    {
+      "epoch": 196.33986928104576,
+      "grad_norm": 0.307283490896225,
+      "learning_rate": 0.00013232916179632193,
+      "loss": 0.1156,
+      "step": 7510
+    },
+    {
+      "epoch": 196.6013071895425,
+      "grad_norm": 0.3378380835056305,
+      "learning_rate": 0.00013217264992678907,
+      "loss": 0.1156,
+      "step": 7520
+    },
+    {
+      "epoch": 196.86274509803923,
+      "grad_norm": 0.3563665449619293,
+      "learning_rate": 0.00013201605009845977,
+      "loss": 0.1197,
+      "step": 7530
+    },
+    {
+      "epoch": 197.12418300653596,
+      "grad_norm": 0.32628506422042847,
+      "learning_rate": 0.00013185936273947207,
+      "loss": 0.118,
+      "step": 7540
+    },
+    {
+      "epoch": 197.3856209150327,
+      "grad_norm": 0.4200800657272339,
+      "learning_rate": 0.00013170258827820318,
+      "loss": 0.1185,
+      "step": 7550
+    },
+    {
+      "epoch": 197.64705882352942,
+      "grad_norm": 0.39217332005500793,
+      "learning_rate": 0.00013154572714326848,
+      "loss": 0.1141,
+      "step": 7560
+    },
+    {
+      "epoch": 197.90849673202615,
+      "grad_norm": 0.37061387300491333,
+      "learning_rate": 0.00013138877976352035,
+      "loss": 0.1179,
+      "step": 7570
+    },
+    {
+      "epoch": 198.16993464052288,
+      "grad_norm": 0.311758816242218,
+      "learning_rate": 0.00013123174656804693,
+      "loss": 0.1191,
+      "step": 7580
+    },
+    {
+      "epoch": 198.4313725490196,
+      "grad_norm": 0.35725489258766174,
+      "learning_rate": 0.00013107462798617097,
+      "loss": 0.1135,
+      "step": 7590
+    },
+    {
+      "epoch": 198.69281045751634,
+      "grad_norm": 0.35550937056541443,
+      "learning_rate": 0.0001309174244474487,
+      "loss": 0.117,
+      "step": 7600
+    },
+    {
+      "epoch": 198.95424836601308,
+      "grad_norm": 0.3795972466468811,
+      "learning_rate": 0.00013076013638166852,
+      "loss": 0.1215,
+      "step": 7610
+    },
+    {
+      "epoch": 199.2156862745098,
+      "grad_norm": 0.3138248026371002,
+      "learning_rate": 0.0001306027642188501,
+      "loss": 0.1133,
+      "step": 7620
+    },
+    {
+      "epoch": 199.47712418300654,
+      "grad_norm": 0.4113026559352875,
+      "learning_rate": 0.00013044530838924283,
+      "loss": 0.1178,
+      "step": 7630
+    },
+    {
+      "epoch": 199.73856209150327,
+      "grad_norm": 0.3749026358127594,
+      "learning_rate": 0.00013028776932332497,
+      "loss": 0.1183,
+      "step": 7640
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 0.4342474043369293,
+      "learning_rate": 0.00013013014745180237,
+      "loss": 0.1204,
+      "step": 7650
+    },
+    {
+      "epoch": 200.26143790849673,
+      "grad_norm": 0.348385214805603,
+      "learning_rate": 0.0001299724432056071,
+      "loss": 0.1152,
+      "step": 7660
+    },
+    {
+      "epoch": 200.52287581699346,
+      "grad_norm": 0.4169091284275055,
+      "learning_rate": 0.00012981465701589664,
+      "loss": 0.1179,
+      "step": 7670
+    },
+    {
+      "epoch": 200.7843137254902,
+      "grad_norm": 0.36309823393821716,
+      "learning_rate": 0.00012965678931405232,
+      "loss": 0.1173,
+      "step": 7680
+    },
+    {
+      "epoch": 201.04575163398692,
+      "grad_norm": 0.3717479705810547,
+      "learning_rate": 0.00012949884053167846,
+      "loss": 0.1168,
+      "step": 7690
+    },
+    {
+      "epoch": 201.30718954248366,
+      "grad_norm": 0.3208405673503876,
+      "learning_rate": 0.00012934081110060105,
+      "loss": 0.1159,
+      "step": 7700
+    },
+    {
+      "epoch": 201.5686274509804,
+      "grad_norm": 0.348286509513855,
+      "learning_rate": 0.00012918270145286642,
+      "loss": 0.1153,
+      "step": 7710
+    },
+    {
+      "epoch": 201.83006535947712,
+      "grad_norm": 0.418100506067276,
+      "learning_rate": 0.00012902451202074038,
+      "loss": 0.1165,
+      "step": 7720
+    },
+    {
+      "epoch": 202.09150326797385,
+      "grad_norm": 0.3307877480983734,
+      "learning_rate": 0.00012886624323670676,
+      "loss": 0.121,
+      "step": 7730
+    },
+    {
+      "epoch": 202.35294117647058,
+      "grad_norm": 0.45300108194351196,
+      "learning_rate": 0.0001287078955334664,
+      "loss": 0.1147,
+      "step": 7740
+    },
+    {
+      "epoch": 202.6143790849673,
+      "grad_norm": 0.37652915716171265,
+      "learning_rate": 0.00012854946934393586,
+      "loss": 0.1165,
+      "step": 7750
+    },
+    {
+      "epoch": 202.87581699346404,
+      "grad_norm": 0.35546672344207764,
+      "learning_rate": 0.0001283909651012463,
+      "loss": 0.1186,
+      "step": 7760
+    },
+    {
+      "epoch": 203.13725490196077,
+      "grad_norm": 0.3252201974391937,
+      "learning_rate": 0.00012823238323874224,
+      "loss": 0.1182,
+      "step": 7770
+    },
+    {
+      "epoch": 203.3986928104575,
+      "grad_norm": 0.36323535442352295,
+      "learning_rate": 0.00012807372418998045,
+      "loss": 0.1143,
+      "step": 7780
+    },
+    {
+      "epoch": 203.66013071895424,
+      "grad_norm": 0.37268149852752686,
+      "learning_rate": 0.00012791498838872874,
+      "loss": 0.1175,
+      "step": 7790
+    },
+    {
+      "epoch": 203.92156862745097,
+      "grad_norm": 0.3726952075958252,
+      "learning_rate": 0.00012775617626896468,
+      "loss": 0.1182,
+      "step": 7800
+    },
+    {
+      "epoch": 204.18300653594773,
+      "grad_norm": 0.34012719988822937,
+      "learning_rate": 0.0001275972882648746,
+      "loss": 0.1178,
+      "step": 7810
+    },
+    {
+      "epoch": 204.44444444444446,
+      "grad_norm": 0.3655106723308563,
+      "learning_rate": 0.0001274383248108522,
+      "loss": 0.1148,
+      "step": 7820
+    },
+    {
+      "epoch": 204.7058823529412,
+      "grad_norm": 0.3548455238342285,
+      "learning_rate": 0.00012727928634149744,
+      "loss": 0.1191,
+      "step": 7830
+    },
+    {
+      "epoch": 204.96732026143792,
+      "grad_norm": 0.3439328670501709,
+      "learning_rate": 0.00012712017329161553,
+      "loss": 0.1176,
+      "step": 7840
+    },
+    {
+      "epoch": 205.22875816993465,
+      "grad_norm": 0.3686840534210205,
+      "learning_rate": 0.00012696098609621542,
+      "loss": 0.1167,
+      "step": 7850
+    },
+    {
+      "epoch": 205.49019607843138,
+      "grad_norm": 0.39468914270401,
+      "learning_rate": 0.00012680172519050883,
+      "loss": 0.1162,
+      "step": 7860
+    },
+    {
+      "epoch": 205.7516339869281,
+      "grad_norm": 0.31318965554237366,
+      "learning_rate": 0.00012664239100990897,
+      "loss": 0.1176,
+      "step": 7870
+    },
+    {
+      "epoch": 206.01307189542484,
+      "grad_norm": 0.3187226355075836,
+      "learning_rate": 0.00012648298399002946,
+      "loss": 0.1186,
+      "step": 7880
+    },
+    {
+      "epoch": 206.27450980392157,
+      "grad_norm": 0.350243479013443,
+      "learning_rate": 0.0001263235045666829,
+      "loss": 0.1142,
+      "step": 7890
+    },
+    {
+      "epoch": 206.5359477124183,
+      "grad_norm": 0.3626524806022644,
+      "learning_rate": 0.00012616395317588007,
+      "loss": 0.1152,
+      "step": 7900
+    },
+    {
+      "epoch": 206.79738562091504,
+      "grad_norm": 0.3574896454811096,
+      "learning_rate": 0.00012600433025382833,
+      "loss": 0.1173,
+      "step": 7910
+    },
+    {
+      "epoch": 207.05882352941177,
+      "grad_norm": 0.3026224374771118,
+      "learning_rate": 0.00012584463623693064,
+      "loss": 0.1206,
+      "step": 7920
+    },
+    {
+      "epoch": 207.3202614379085,
+      "grad_norm": 0.3898176848888397,
+      "learning_rate": 0.00012568487156178434,
+      "loss": 0.1132,
+      "step": 7930
+    },
+    {
+      "epoch": 207.58169934640523,
+      "grad_norm": 0.3034379780292511,
+      "learning_rate": 0.00012552503666517998,
+      "loss": 0.1162,
+      "step": 7940
+    },
+    {
+      "epoch": 207.84313725490196,
+      "grad_norm": 0.38479849696159363,
+      "learning_rate": 0.00012536513198410006,
+      "loss": 0.1191,
+      "step": 7950
+    },
+    {
+      "epoch": 208.1045751633987,
+      "grad_norm": 0.4212307631969452,
+      "learning_rate": 0.00012520515795571785,
+      "loss": 0.1182,
+      "step": 7960
+    },
+    {
+      "epoch": 208.36601307189542,
+      "grad_norm": 0.44150879979133606,
+      "learning_rate": 0.00012504511501739622,
+      "loss": 0.1161,
+      "step": 7970
+    },
+    {
+      "epoch": 208.62745098039215,
+      "grad_norm": 0.49862542748451233,
+      "learning_rate": 0.0001248850036066865,
+      "loss": 0.1186,
+      "step": 7980
+    },
+    {
+      "epoch": 208.88888888888889,
+      "grad_norm": 0.37723541259765625,
+      "learning_rate": 0.00012472482416132712,
+      "loss": 0.1159,
+      "step": 7990
+    },
+    {
+      "epoch": 209.15032679738562,
+      "grad_norm": 0.39432400465011597,
+      "learning_rate": 0.00012456457711924266,
+      "loss": 0.1139,
+      "step": 8000
+    },
+    {
+      "epoch": 209.41176470588235,
+      "grad_norm": 0.320744127035141,
+      "learning_rate": 0.0001244042629185423,
+      "loss": 0.1138,
+      "step": 8010
+    },
+    {
+      "epoch": 209.67320261437908,
+      "grad_norm": 0.37809300422668457,
+      "learning_rate": 0.00012424388199751903,
+      "loss": 0.1164,
+      "step": 8020
+    },
+    {
+      "epoch": 209.9346405228758,
+      "grad_norm": 0.3541184961795807,
+      "learning_rate": 0.0001240834347946481,
+      "loss": 0.1208,
+      "step": 8030
+    },
+    {
+      "epoch": 210.19607843137254,
+      "grad_norm": 0.38199830055236816,
+      "learning_rate": 0.00012392292174858606,
+      "loss": 0.1171,
+      "step": 8040
+    },
+    {
+      "epoch": 210.45751633986927,
+      "grad_norm": 0.30296802520751953,
+      "learning_rate": 0.00012376234329816949,
+      "loss": 0.1173,
+      "step": 8050
+    },
+    {
+      "epoch": 210.718954248366,
+      "grad_norm": 0.38834238052368164,
+      "learning_rate": 0.00012360169988241367,
+      "loss": 0.1184,
+      "step": 8060
+    },
+    {
+      "epoch": 210.98039215686273,
+      "grad_norm": 0.3523692190647125,
+      "learning_rate": 0.0001234409919405116,
+      "loss": 0.1163,
+      "step": 8070
+    },
+    {
+      "epoch": 211.24183006535947,
+      "grad_norm": 0.31442514061927795,
+      "learning_rate": 0.0001232802199118327,
+      "loss": 0.1154,
+      "step": 8080
+    },
+    {
+      "epoch": 211.5032679738562,
+      "grad_norm": 0.365710586309433,
+      "learning_rate": 0.00012311938423592152,
+      "loss": 0.1143,
+      "step": 8090
+    },
+    {
+      "epoch": 211.76470588235293,
+      "grad_norm": 0.36913615465164185,
+      "learning_rate": 0.00012295848535249658,
+      "loss": 0.1169,
+      "step": 8100
+    },
+    {
+      "epoch": 212.0261437908497,
+      "grad_norm": 0.2978919744491577,
+      "learning_rate": 0.0001227975237014494,
+      "loss": 0.1208,
+      "step": 8110
+    },
+    {
+      "epoch": 212.28758169934642,
+      "grad_norm": 0.3304121494293213,
+      "learning_rate": 0.00012263649972284294,
+      "loss": 0.1137,
+      "step": 8120
+    },
+    {
+      "epoch": 212.54901960784315,
+      "grad_norm": 0.3513433337211609,
+      "learning_rate": 0.00012247541385691058,
+      "loss": 0.1198,
+      "step": 8130
+    },
+    {
+      "epoch": 212.81045751633988,
+      "grad_norm": 0.4109828472137451,
+      "learning_rate": 0.000122314266544055,
+      "loss": 0.1185,
+      "step": 8140
+    },
+    {
+      "epoch": 213.0718954248366,
+      "grad_norm": 0.2574869692325592,
+      "learning_rate": 0.00012215305822484672,
+      "loss": 0.1133,
+      "step": 8150
+    },
+    {
+      "epoch": 213.33333333333334,
+      "grad_norm": 0.307307630777359,
+      "learning_rate": 0.00012199178934002317,
+      "loss": 0.1153,
+      "step": 8160
+    },
+    {
+      "epoch": 213.59477124183007,
+      "grad_norm": 0.3619244396686554,
+      "learning_rate": 0.00012183046033048736,
+      "loss": 0.1173,
+      "step": 8170
+    },
+    {
+      "epoch": 213.8562091503268,
+      "grad_norm": 0.39120060205459595,
+      "learning_rate": 0.00012166907163730656,
+      "loss": 0.1166,
+      "step": 8180
+    },
+    {
+      "epoch": 214.11764705882354,
+      "grad_norm": 0.3365266025066376,
+      "learning_rate": 0.00012150762370171136,
+      "loss": 0.1201,
+      "step": 8190
+    },
+    {
+      "epoch": 214.37908496732027,
+      "grad_norm": 0.4538247883319855,
+      "learning_rate": 0.00012134611696509419,
+      "loss": 0.1157,
+      "step": 8200
+    },
+    {
+      "epoch": 214.640522875817,
+      "grad_norm": 0.3183276951313019,
+      "learning_rate": 0.00012118455186900836,
+      "loss": 0.1128,
+      "step": 8210
+    },
+    {
+      "epoch": 214.90196078431373,
+      "grad_norm": 0.8887193202972412,
+      "learning_rate": 0.00012102292885516666,
+      "loss": 0.1214,
+      "step": 8220
+    },
+    {
+      "epoch": 215.16339869281046,
+      "grad_norm": 0.3512011766433716,
+      "learning_rate": 0.00012086124836544024,
+      "loss": 0.1171,
+      "step": 8230
+    },
+    {
+      "epoch": 215.4248366013072,
+      "grad_norm": 0.3707999289035797,
+      "learning_rate": 0.00012069951084185733,
+      "loss": 0.1164,
+      "step": 8240
+    },
+    {
+      "epoch": 215.68627450980392,
+      "grad_norm": 0.3448478877544403,
+      "learning_rate": 0.00012053771672660221,
+      "loss": 0.1188,
+      "step": 8250
+    },
+    {
+      "epoch": 215.94771241830065,
+      "grad_norm": 0.3120673596858978,
+      "learning_rate": 0.00012037586646201378,
+      "loss": 0.1177,
+      "step": 8260
+    },
+    {
+      "epoch": 216.20915032679738,
+      "grad_norm": 0.46193230152130127,
+      "learning_rate": 0.00012021396049058451,
+      "loss": 0.1163,
+      "step": 8270
+    },
+    {
+      "epoch": 216.47058823529412,
+      "grad_norm": 0.34690043330192566,
+      "learning_rate": 0.00012005199925495914,
+      "loss": 0.1157,
+      "step": 8280
+    },
+    {
+      "epoch": 216.73202614379085,
+      "grad_norm": 0.4091133177280426,
+      "learning_rate": 0.00011988998319793346,
+      "loss": 0.1175,
+      "step": 8290
+    },
+    {
+      "epoch": 216.99346405228758,
+      "grad_norm": 0.32104718685150146,
+      "learning_rate": 0.00011972791276245321,
+      "loss": 0.1182,
+      "step": 8300
+    },
+    {
+      "epoch": 217.2549019607843,
+      "grad_norm": 0.3396114408969879,
+      "learning_rate": 0.00011956578839161279,
+      "loss": 0.1178,
+      "step": 8310
+    },
+    {
+      "epoch": 217.51633986928104,
+      "grad_norm": 0.39414018392562866,
+      "learning_rate": 0.00011940361052865401,
+      "loss": 0.1184,
+      "step": 8320
+    },
+    {
+      "epoch": 217.77777777777777,
+      "grad_norm": 0.3328598141670227,
+      "learning_rate": 0.000119241379616965,
+      "loss": 0.114,
+      "step": 8330
+    },
+    {
+      "epoch": 218.0392156862745,
+      "grad_norm": 0.334634929895401,
+      "learning_rate": 0.00011907909610007884,
+      "loss": 0.116,
+      "step": 8340
+    },
+    {
+      "epoch": 218.30065359477123,
+      "grad_norm": 0.32335221767425537,
+      "learning_rate": 0.00011891676042167246,
+      "loss": 0.1176,
+      "step": 8350
+    },
+    {
+      "epoch": 218.56209150326796,
+      "grad_norm": 0.2827315926551819,
+      "learning_rate": 0.00011875437302556543,
+      "loss": 0.1162,
+      "step": 8360
+    },
+    {
+      "epoch": 218.8235294117647,
+      "grad_norm": 0.3528691828250885,
+      "learning_rate": 0.0001185919343557187,
+      "loss": 0.1164,
+      "step": 8370
+    },
+    {
+      "epoch": 219.08496732026143,
+      "grad_norm": 0.28620168566703796,
+      "learning_rate": 0.00011842944485623335,
+      "loss": 0.1172,
+      "step": 8380
+    },
+    {
+      "epoch": 219.34640522875816,
+      "grad_norm": 0.38785773515701294,
+      "learning_rate": 0.0001182669049713495,
+      "loss": 0.1156,
+      "step": 8390
+    },
+    {
+      "epoch": 219.6078431372549,
+      "grad_norm": 0.30017638206481934,
+      "learning_rate": 0.00011810431514544496,
+      "loss": 0.1149,
+      "step": 8400
+    },
+    {
+      "epoch": 219.86928104575162,
+      "grad_norm": 0.3550892770290375,
+      "learning_rate": 0.00011794167582303412,
+      "loss": 0.1184,
+      "step": 8410
+    },
+    {
+      "epoch": 220.13071895424838,
+      "grad_norm": 0.2966924011707306,
+      "learning_rate": 0.00011777898744876673,
+      "loss": 0.1166,
+      "step": 8420
+    },
+    {
+      "epoch": 220.3921568627451,
+      "grad_norm": 0.2973492443561554,
+      "learning_rate": 0.00011761625046742651,
+      "loss": 0.1152,
+      "step": 8430
+    },
+    {
+      "epoch": 220.65359477124184,
+      "grad_norm": 0.3460974395275116,
+      "learning_rate": 0.00011745346532393017,
+      "loss": 0.115,
+      "step": 8440
+    },
+    {
+      "epoch": 220.91503267973857,
+      "grad_norm": 0.3842596113681793,
+      "learning_rate": 0.0001172906324633261,
+      "loss": 0.118,
+      "step": 8450
+    },
+    {
+      "epoch": 221.1764705882353,
+      "grad_norm": 0.3122817575931549,
+      "learning_rate": 0.00011712775233079311,
+      "loss": 0.1159,
+      "step": 8460
+    },
+    {
+      "epoch": 221.43790849673204,
+      "grad_norm": 0.3846839368343353,
+      "learning_rate": 0.00011696482537163933,
+      "loss": 0.1144,
+      "step": 8470
+    },
+    {
+      "epoch": 221.69934640522877,
+      "grad_norm": 0.3940717875957489,
+      "learning_rate": 0.00011680185203130075,
+      "loss": 0.1156,
+      "step": 8480
+    },
+    {
+      "epoch": 221.9607843137255,
+      "grad_norm": 0.35572728514671326,
+      "learning_rate": 0.00011663883275534029,
+      "loss": 0.1221,
+      "step": 8490
+    },
+    {
+      "epoch": 222.22222222222223,
+      "grad_norm": 0.2821555733680725,
+      "learning_rate": 0.0001164757679894464,
+      "loss": 0.1159,
+      "step": 8500
+    },
+    {
+      "epoch": 222.48366013071896,
+      "grad_norm": 0.4151257574558258,
+      "learning_rate": 0.00011631265817943198,
+      "loss": 0.1136,
+      "step": 8510
+    },
+    {
+      "epoch": 222.7450980392157,
+      "grad_norm": 0.3188461661338806,
+      "learning_rate": 0.000116149503771233,
+      "loss": 0.1156,
+      "step": 8520
+    },
+    {
+      "epoch": 223.00653594771242,
+      "grad_norm": 0.3545646667480469,
+      "learning_rate": 0.00011598630521090734,
+      "loss": 0.1183,
+      "step": 8530
+    },
+    {
+      "epoch": 223.26797385620915,
+      "grad_norm": 0.3330908417701721,
+      "learning_rate": 0.00011582306294463372,
+      "loss": 0.1144,
+      "step": 8540
+    },
+    {
+      "epoch": 223.52941176470588,
+      "grad_norm": 0.3198162913322449,
+      "learning_rate": 0.00011565977741871018,
+      "loss": 0.1182,
+      "step": 8550
+    },
+    {
+      "epoch": 223.79084967320262,
+      "grad_norm": 0.38695403933525085,
+      "learning_rate": 0.00011549644907955315,
+      "loss": 0.1156,
+      "step": 8560
+    },
+    {
+      "epoch": 224.05228758169935,
+      "grad_norm": 0.3442593514919281,
+      "learning_rate": 0.00011533307837369607,
+      "loss": 0.1191,
+      "step": 8570
+    },
+    {
+      "epoch": 224.31372549019608,
+      "grad_norm": 0.3656689524650574,
+      "learning_rate": 0.00011516966574778822,
+      "loss": 0.1154,
+      "step": 8580
+    },
+    {
+      "epoch": 224.5751633986928,
+      "grad_norm": 0.3869684636592865,
+      "learning_rate": 0.00011500621164859347,
+      "loss": 0.1176,
+      "step": 8590
+    },
+    {
+      "epoch": 224.83660130718954,
+      "grad_norm": 0.36732858419418335,
+      "learning_rate": 0.00011484271652298906,
+      "loss": 0.1121,
+      "step": 8600
+    },
+    {
+      "epoch": 225.09803921568627,
+      "grad_norm": 0.35245758295059204,
+      "learning_rate": 0.00011467918081796445,
+      "loss": 0.1189,
+      "step": 8610
+    },
+    {
+      "epoch": 225.359477124183,
+      "grad_norm": 0.3855646848678589,
+      "learning_rate": 0.00011451560498062,
+      "loss": 0.1169,
+      "step": 8620
+    },
+    {
+      "epoch": 225.62091503267973,
+      "grad_norm": 0.28450724482536316,
+      "learning_rate": 0.00011435198945816584,
+      "loss": 0.1142,
+      "step": 8630
+    },
+    {
+      "epoch": 225.88235294117646,
+      "grad_norm": 0.3665756583213806,
+      "learning_rate": 0.00011418833469792047,
+      "loss": 0.1202,
+      "step": 8640
+    },
+    {
+      "epoch": 226.1437908496732,
+      "grad_norm": 0.23904922604560852,
+      "learning_rate": 0.00011402464114730989,
+      "loss": 0.1133,
+      "step": 8650
+    },
+    {
+      "epoch": 226.40522875816993,
+      "grad_norm": 0.3275870382785797,
+      "learning_rate": 0.0001138609092538659,
+      "loss": 0.1138,
+      "step": 8660
+    },
+    {
+      "epoch": 226.66666666666666,
+      "grad_norm": 0.3351685404777527,
+      "learning_rate": 0.00011369713946522532,
+      "loss": 0.1143,
+      "step": 8670
+    },
+    {
+      "epoch": 226.9281045751634,
+      "grad_norm": 0.29419705271720886,
+      "learning_rate": 0.00011353333222912843,
+      "loss": 0.1198,
+      "step": 8680
+    },
+    {
+      "epoch": 227.18954248366012,
+      "grad_norm": 0.3791927993297577,
+      "learning_rate": 0.00011336948799341798,
+      "loss": 0.116,
+      "step": 8690
+    },
+    {
+      "epoch": 227.45098039215685,
+      "grad_norm": 0.487420916557312,
+      "learning_rate": 0.00011320560720603792,
+      "loss": 0.1183,
+      "step": 8700
+    },
+    {
+      "epoch": 227.71241830065358,
+      "grad_norm": 0.2873092591762543,
+      "learning_rate": 0.00011304169031503197,
+      "loss": 0.1146,
+      "step": 8710
+    },
+    {
+      "epoch": 227.9738562091503,
+      "grad_norm": 0.3828370273113251,
+      "learning_rate": 0.00011287773776854273,
+      "loss": 0.1202,
+      "step": 8720
+    },
+    {
+      "epoch": 228.23529411764707,
+      "grad_norm": 0.3244374990463257,
+      "learning_rate": 0.00011271375001481015,
+      "loss": 0.1141,
+      "step": 8730
+    },
+    {
+      "epoch": 228.4967320261438,
+      "grad_norm": 0.3474271893501282,
+      "learning_rate": 0.0001125497275021705,
+      "loss": 0.1144,
+      "step": 8740
+    },
+    {
+      "epoch": 228.75816993464053,
+      "grad_norm": 0.43531376123428345,
+      "learning_rate": 0.00011238567067905507,
+      "loss": 0.1183,
+      "step": 8750
+    },
+    {
+      "epoch": 229.01960784313727,
+      "grad_norm": 0.27749860286712646,
+      "learning_rate": 0.00011222157999398895,
+      "loss": 0.1166,
+      "step": 8760
+    },
+    {
+      "epoch": 229.281045751634,
+      "grad_norm": 0.35005658864974976,
+      "learning_rate": 0.00011205745589558982,
+      "loss": 0.1135,
+      "step": 8770
+    },
+    {
+      "epoch": 229.54248366013073,
+      "grad_norm": 0.3147236704826355,
+      "learning_rate": 0.00011189329883256668,
+      "loss": 0.1161,
+      "step": 8780
+    },
+    {
+      "epoch": 229.80392156862746,
+      "grad_norm": 0.3600068688392639,
+      "learning_rate": 0.00011172910925371865,
+      "loss": 0.121,
+      "step": 8790
+    },
+    {
+      "epoch": 230.0653594771242,
+      "grad_norm": 0.2795291543006897,
+      "learning_rate": 0.00011156488760793383,
+      "loss": 0.1145,
+      "step": 8800
+    },
+    {
+      "epoch": 230.32679738562092,
+      "grad_norm": 0.33064013719558716,
+      "learning_rate": 0.00011140063434418788,
+      "loss": 0.1133,
+      "step": 8810
+    },
+    {
+      "epoch": 230.58823529411765,
+      "grad_norm": 0.38541939854621887,
+      "learning_rate": 0.00011123634991154294,
+      "loss": 0.1172,
+      "step": 8820
+    },
+    {
+      "epoch": 230.84967320261438,
+      "grad_norm": 0.4045771062374115,
+      "learning_rate": 0.00011107203475914643,
+      "loss": 0.1183,
+      "step": 8830
+    },
+    {
+      "epoch": 231.11111111111111,
+      "grad_norm": 0.2556995153427124,
+      "learning_rate": 0.00011090768933622966,
+      "loss": 0.1141,
+      "step": 8840
+    },
+    {
+      "epoch": 231.37254901960785,
+      "grad_norm": 0.2857918441295624,
+      "learning_rate": 0.00011074331409210677,
+      "loss": 0.1128,
+      "step": 8850
+    },
+    {
+      "epoch": 231.63398692810458,
+      "grad_norm": 0.4650552570819855,
+      "learning_rate": 0.00011057890947617338,
+      "loss": 0.1198,
+      "step": 8860
+    },
+    {
+      "epoch": 231.8954248366013,
+      "grad_norm": 0.31443580985069275,
+      "learning_rate": 0.00011041447593790544,
+      "loss": 0.1166,
+      "step": 8870
+    },
+    {
+      "epoch": 232.15686274509804,
+      "grad_norm": 0.32467028498649597,
+      "learning_rate": 0.00011025001392685794,
+      "loss": 0.1167,
+      "step": 8880
+    },
+    {
+      "epoch": 232.41830065359477,
+      "grad_norm": 0.2936582565307617,
+      "learning_rate": 0.00011008552389266376,
+      "loss": 0.1145,
+      "step": 8890
+    },
+    {
+      "epoch": 232.6797385620915,
+      "grad_norm": 0.31260111927986145,
+      "learning_rate": 0.0001099210062850324,
+      "loss": 0.1147,
+      "step": 8900
+    },
+    {
+      "epoch": 232.94117647058823,
+      "grad_norm": 0.2891233265399933,
+      "learning_rate": 0.00010975646155374868,
+      "loss": 0.1194,
+      "step": 8910
+    },
+    {
+      "epoch": 233.20261437908496,
+      "grad_norm": 0.3409523367881775,
+      "learning_rate": 0.00010959189014867161,
+      "loss": 0.1156,
+      "step": 8920
+    },
+    {
+      "epoch": 233.4640522875817,
+      "grad_norm": 0.29765674471855164,
+      "learning_rate": 0.00010942729251973313,
+      "loss": 0.1151,
+      "step": 8930
+    },
+    {
+      "epoch": 233.72549019607843,
+      "grad_norm": 0.3181908428668976,
+      "learning_rate": 0.0001092626691169369,
+      "loss": 0.1181,
+      "step": 8940
+    },
+    {
+      "epoch": 233.98692810457516,
+      "grad_norm": 0.3303356170654297,
+      "learning_rate": 0.00010909802039035701,
+      "loss": 0.1163,
+      "step": 8950
+    },
+    {
+      "epoch": 234.2483660130719,
+      "grad_norm": 0.3642708957195282,
+      "learning_rate": 0.00010893334679013676,
+      "loss": 0.1132,
+      "step": 8960
+    },
+    {
+      "epoch": 234.50980392156862,
+      "grad_norm": 0.36728131771087646,
+      "learning_rate": 0.00010876864876648751,
+      "loss": 0.1159,
+      "step": 8970
+    },
+    {
+      "epoch": 234.77124183006535,
+      "grad_norm": 0.34145310521125793,
+      "learning_rate": 0.00010860392676968736,
+      "loss": 0.1168,
+      "step": 8980
+    },
+    {
+      "epoch": 235.03267973856208,
+      "grad_norm": 0.34180009365081787,
+      "learning_rate": 0.00010843918125007995,
+      "loss": 0.1201,
+      "step": 8990
+    },
+    {
+      "epoch": 235.2941176470588,
+      "grad_norm": 0.3454407751560211,
+      "learning_rate": 0.00010827441265807328,
+      "loss": 0.1149,
+      "step": 9000
+    },
+    {
+      "epoch": 235.55555555555554,
+      "grad_norm": 0.3704677224159241,
+      "learning_rate": 0.00010810962144413834,
+      "loss": 0.1117,
+      "step": 9010
+    },
+    {
+      "epoch": 235.81699346405227,
+      "grad_norm": 0.3329799175262451,
+      "learning_rate": 0.00010794480805880804,
+      "loss": 0.1181,
+      "step": 9020
+    },
+    {
+      "epoch": 236.07843137254903,
+      "grad_norm": 0.3676550090312958,
+      "learning_rate": 0.00010777997295267588,
+      "loss": 0.1148,
+      "step": 9030
+    },
+    {
+      "epoch": 236.33986928104576,
+      "grad_norm": 0.34352195262908936,
+      "learning_rate": 0.00010761511657639474,
+      "loss": 0.1154,
+      "step": 9040
+    },
+    {
+      "epoch": 236.6013071895425,
+      "grad_norm": 0.3431718349456787,
+      "learning_rate": 0.00010745023938067569,
+      "loss": 0.1154,
+      "step": 9050
+    },
+    {
+      "epoch": 236.86274509803923,
+      "grad_norm": 0.37354880571365356,
+      "learning_rate": 0.00010728534181628665,
+      "loss": 0.1186,
+      "step": 9060
+    },
+    {
+      "epoch": 237.12418300653596,
+      "grad_norm": 0.2959931790828705,
+      "learning_rate": 0.00010712042433405125,
+      "loss": 0.1153,
+      "step": 9070
+    },
+    {
+      "epoch": 237.3856209150327,
+      "grad_norm": 0.3524629771709442,
+      "learning_rate": 0.00010695548738484762,
+      "loss": 0.1135,
+      "step": 9080
+    },
+    {
+      "epoch": 237.64705882352942,
+      "grad_norm": 0.37540528178215027,
+      "learning_rate": 0.00010679053141960709,
+      "loss": 0.1199,
+      "step": 9090
+    },
+    {
+      "epoch": 237.90849673202615,
+      "grad_norm": 0.4271695017814636,
+      "learning_rate": 0.00010662555688931295,
+      "loss": 0.119,
+      "step": 9100
+    },
+    {
+      "epoch": 238.16993464052288,
+      "grad_norm": 0.3042963147163391,
+      "learning_rate": 0.00010646056424499926,
+      "loss": 0.1134,
+      "step": 9110
+    },
+    {
+      "epoch": 238.4313725490196,
+      "grad_norm": 0.3255995512008667,
+      "learning_rate": 0.00010629555393774962,
+      "loss": 0.1155,
+      "step": 9120
+    },
+    {
+      "epoch": 238.69281045751634,
+      "grad_norm": 0.3537774980068207,
+      "learning_rate": 0.00010613052641869592,
+      "loss": 0.1181,
+      "step": 9130
+    },
+    {
+      "epoch": 238.95424836601308,
+      "grad_norm": 0.34757131338119507,
+      "learning_rate": 0.00010596548213901708,
+      "loss": 0.1167,
+      "step": 9140
+    },
+    {
+      "epoch": 239.2156862745098,
+      "grad_norm": 0.2638850808143616,
+      "learning_rate": 0.00010580042154993786,
+      "loss": 0.1128,
+      "step": 9150
+    },
+    {
+      "epoch": 239.47712418300654,
+      "grad_norm": 0.39982908964157104,
+      "learning_rate": 0.00010563534510272763,
+      "loss": 0.1166,
+      "step": 9160
+    },
+    {
+      "epoch": 239.73856209150327,
+      "grad_norm": 0.4114007353782654,
+      "learning_rate": 0.00010547025324869903,
+      "loss": 0.1164,
+      "step": 9170
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 0.4774865210056305,
+      "learning_rate": 0.00010530514643920697,
+      "loss": 0.1184,
+      "step": 9180
+    },
+    {
+      "epoch": 240.26143790849673,
+      "grad_norm": 0.2834888994693756,
+      "learning_rate": 0.00010514002512564714,
+      "loss": 0.112,
+      "step": 9190
+    },
+    {
+      "epoch": 240.52287581699346,
+      "grad_norm": 0.3496600091457367,
+      "learning_rate": 0.00010497488975945484,
+      "loss": 0.1152,
+      "step": 9200
+    },
+    {
+      "epoch": 240.7843137254902,
+      "grad_norm": 0.35766535997390747,
+      "learning_rate": 0.00010480974079210392,
+      "loss": 0.1181,
+      "step": 9210
+    },
+    {
+      "epoch": 241.04575163398692,
+      "grad_norm": 0.30494368076324463,
+      "learning_rate": 0.00010464457867510533,
+      "loss": 0.1177,
+      "step": 9220
+    },
+    {
+      "epoch": 241.30718954248366,
+      "grad_norm": 0.33818888664245605,
+      "learning_rate": 0.00010447940386000601,
+      "loss": 0.1138,
+      "step": 9230
+    },
+    {
+      "epoch": 241.5686274509804,
+      "grad_norm": 0.36898425221443176,
+      "learning_rate": 0.00010431421679838758,
+      "loss": 0.1136,
+      "step": 9240
+    },
+    {
+      "epoch": 241.83006535947712,
+      "grad_norm": 0.3632810711860657,
+      "learning_rate": 0.00010414901794186514,
+      "loss": 0.1196,
+      "step": 9250
+    },
+    {
+      "epoch": 242.09150326797385,
+      "grad_norm": 0.3109024167060852,
+      "learning_rate": 0.00010398380774208609,
+      "loss": 0.1166,
+      "step": 9260
+    },
+    {
+      "epoch": 242.35294117647058,
+      "grad_norm": 0.29130157828330994,
+      "learning_rate": 0.00010381858665072878,
+      "loss": 0.1143,
+      "step": 9270
+    },
+    {
+      "epoch": 242.6143790849673,
+      "grad_norm": 0.32316336035728455,
+      "learning_rate": 0.00010365335511950142,
+      "loss": 0.1122,
+      "step": 9280
+    },
+    {
+      "epoch": 242.87581699346404,
+      "grad_norm": 0.3277498185634613,
+      "learning_rate": 0.00010348811360014063,
+      "loss": 0.1192,
+      "step": 9290
+    },
+    {
+      "epoch": 243.13725490196077,
+      "grad_norm": 0.3015691637992859,
+      "learning_rate": 0.00010332286254441049,
+      "loss": 0.1161,
+      "step": 9300
+    },
+    {
+      "epoch": 243.3986928104575,
+      "grad_norm": 0.36535218358039856,
+      "learning_rate": 0.00010315760240410097,
+      "loss": 0.1162,
+      "step": 9310
+    },
+    {
+      "epoch": 243.66013071895424,
+      "grad_norm": 0.40011629462242126,
+      "learning_rate": 0.0001029923336310271,
+      "loss": 0.1183,
+      "step": 9320
+    },
+    {
+      "epoch": 243.92156862745097,
+      "grad_norm": 0.3597753643989563,
+      "learning_rate": 0.00010282705667702734,
+      "loss": 0.115,
+      "step": 9330
+    },
+    {
+      "epoch": 244.18300653594773,
+      "grad_norm": 0.37653326988220215,
+      "learning_rate": 0.00010266177199396257,
+      "loss": 0.1148,
+      "step": 9340
+    },
+    {
+      "epoch": 244.44444444444446,
+      "grad_norm": 0.30488893389701843,
+      "learning_rate": 0.00010249648003371482,
+      "loss": 0.1152,
+      "step": 9350
+    },
+    {
+      "epoch": 244.7058823529412,
+      "grad_norm": 0.351938396692276,
+      "learning_rate": 0.00010233118124818595,
+      "loss": 0.1174,
+      "step": 9360
+    },
+    {
+      "epoch": 244.96732026143792,
+      "grad_norm": 0.3121070861816406,
+      "learning_rate": 0.0001021658760892966,
+      "loss": 0.1136,
+      "step": 9370
+    },
+    {
+      "epoch": 245.22875816993465,
+      "grad_norm": 0.33020684123039246,
+      "learning_rate": 0.0001020005650089847,
+      "loss": 0.1108,
+      "step": 9380
+    },
+    {
+      "epoch": 245.49019607843138,
+      "grad_norm": 0.431669145822525,
+      "learning_rate": 0.00010183524845920447,
+      "loss": 0.1175,
+      "step": 9390
+    },
+    {
+      "epoch": 245.7516339869281,
+      "grad_norm": 0.2955373227596283,
+      "learning_rate": 0.00010166992689192505,
+      "loss": 0.1176,
+      "step": 9400
+    },
+    {
+      "epoch": 246.01307189542484,
+      "grad_norm": 0.3435670733451843,
+      "learning_rate": 0.00010150460075912922,
+      "loss": 0.1167,
+      "step": 9410
+    },
+    {
+      "epoch": 246.27450980392157,
+      "grad_norm": 0.3121711313724518,
+      "learning_rate": 0.00010133927051281243,
+      "loss": 0.114,
+      "step": 9420
+    },
+    {
+      "epoch": 246.5359477124183,
+      "grad_norm": 0.32946914434432983,
+      "learning_rate": 0.00010117393660498116,
+      "loss": 0.1133,
+      "step": 9430
+    },
+    {
+      "epoch": 246.79738562091504,
+      "grad_norm": 0.30625325441360474,
+      "learning_rate": 0.00010100859948765204,
+      "loss": 0.119,
+      "step": 9440
+    },
+    {
+      "epoch": 247.05882352941177,
+      "grad_norm": 0.312551885843277,
+      "learning_rate": 0.00010084325961285046,
+      "loss": 0.1157,
+      "step": 9450
+    },
+    {
+      "epoch": 247.3202614379085,
+      "grad_norm": 0.24457718431949615,
+      "learning_rate": 0.00010067791743260924,
+      "loss": 0.11,
+      "step": 9460
+    },
+    {
+      "epoch": 247.58169934640523,
+      "grad_norm": 0.33879950642585754,
+      "learning_rate": 0.00010051257339896771,
+      "loss": 0.1174,
+      "step": 9470
+    },
+    {
+      "epoch": 247.84313725490196,
+      "grad_norm": 0.33240655064582825,
+      "learning_rate": 0.00010034722796397004,
+      "loss": 0.1179,
+      "step": 9480
+    },
+    {
+      "epoch": 248.1045751633987,
+      "grad_norm": 0.30761128664016724,
+      "learning_rate": 0.00010018188157966442,
+      "loss": 0.1172,
+      "step": 9490
+    },
+    {
+      "epoch": 248.36601307189542,
+      "grad_norm": 0.30122920870780945,
+      "learning_rate": 0.00010001653469810145,
+      "loss": 0.1152,
+      "step": 9500
+    },
+    {
+      "epoch": 248.62745098039215,
+      "grad_norm": 0.3038076162338257,
+      "learning_rate": 9.985118777133329e-05,
+      "loss": 0.1161,
+      "step": 9510
+    },
+    {
+      "epoch": 248.88888888888889,
+      "grad_norm": 0.38542667031288147,
+      "learning_rate": 9.968584125141204e-05,
+      "loss": 0.1167,
+      "step": 9520
+    },
+    {
+      "epoch": 249.15032679738562,
+      "grad_norm": 0.2926845848560333,
+      "learning_rate": 9.952049559038885e-05,
+      "loss": 0.1153,
+      "step": 9530
+    },
+    {
+      "epoch": 249.41176470588235,
+      "grad_norm": 0.35238757729530334,
+      "learning_rate": 9.935515124031239e-05,
+      "loss": 0.1134,
+      "step": 9540
+    },
+    {
+      "epoch": 249.67320261437908,
+      "grad_norm": 0.31012189388275146,
+      "learning_rate": 9.918980865322782e-05,
+      "loss": 0.116,
+      "step": 9550
+    },
+    {
+      "epoch": 249.9346405228758,
+      "grad_norm": 0.3544747531414032,
+      "learning_rate": 9.902446828117545e-05,
+      "loss": 0.1189,
+      "step": 9560
+    },
+    {
+      "epoch": 250.19607843137254,
+      "grad_norm": 0.4231995940208435,
+      "learning_rate": 9.885913057618955e-05,
+      "loss": 0.1148,
+      "step": 9570
+    },
+    {
+      "epoch": 250.45751633986927,
+      "grad_norm": 0.3783901333808899,
+      "learning_rate": 9.869379599029708e-05,
+      "loss": 0.1158,
+      "step": 9580
+    },
+    {
+      "epoch": 250.718954248366,
+      "grad_norm": 0.32323822379112244,
+      "learning_rate": 9.852846497551653e-05,
+      "loss": 0.1172,
+      "step": 9590
+    },
+    {
+      "epoch": 250.98039215686273,
+      "grad_norm": 0.34913671016693115,
+      "learning_rate": 9.836313798385653e-05,
+      "loss": 0.1143,
+      "step": 9600
+    },
+    {
+      "epoch": 251.24183006535947,
+      "grad_norm": 0.2595481276512146,
+      "learning_rate": 9.819781546731476e-05,
+      "loss": 0.1137,
+      "step": 9610
+    },
+    {
+      "epoch": 251.5032679738562,
+      "grad_norm": 0.30071359872817993,
+      "learning_rate": 9.803249787787669e-05,
+      "loss": 0.1141,
+      "step": 9620
+    },
+    {
+      "epoch": 251.76470588235293,
+      "grad_norm": 0.32096344232559204,
+      "learning_rate": 9.786718566751431e-05,
+      "loss": 0.1176,
+      "step": 9630
+    },
+    {
+      "epoch": 252.0261437908497,
+      "grad_norm": 0.30524900555610657,
+      "learning_rate": 9.770187928818487e-05,
+      "loss": 0.118,
+      "step": 9640
+    },
+    {
+      "epoch": 252.28758169934642,
+      "grad_norm": 0.3532697856426239,
+      "learning_rate": 9.75365791918297e-05,
+      "loss": 0.1158,
+      "step": 9650
+    },
+    {
+      "epoch": 252.54901960784315,
+      "grad_norm": 0.29913774132728577,
+      "learning_rate": 9.737128583037295e-05,
+      "loss": 0.1145,
+      "step": 9660
+    },
+    {
+      "epoch": 252.81045751633988,
+      "grad_norm": 0.35638126730918884,
+      "learning_rate": 9.720599965572036e-05,
+      "loss": 0.1145,
+      "step": 9670
+    },
+    {
+      "epoch": 253.0718954248366,
+      "grad_norm": 0.28130611777305603,
+      "learning_rate": 9.704072111975802e-05,
+      "loss": 0.1143,
+      "step": 9680
+    },
+    {
+      "epoch": 253.33333333333334,
+      "grad_norm": 0.39420101046562195,
+      "learning_rate": 9.687545067435116e-05,
+      "loss": 0.1128,
+      "step": 9690
+    },
+    {
+      "epoch": 253.59477124183007,
+      "grad_norm": 0.33940833806991577,
+      "learning_rate": 9.671018877134284e-05,
+      "loss": 0.1152,
+      "step": 9700
+    },
+    {
+      "epoch": 253.8562091503268,
+      "grad_norm": 0.38873952627182007,
+      "learning_rate": 9.654493586255278e-05,
+      "loss": 0.1207,
+      "step": 9710
+    },
+    {
+      "epoch": 254.11764705882354,
+      "grad_norm": 0.242060124874115,
+      "learning_rate": 9.637969239977614e-05,
+      "loss": 0.1143,
+      "step": 9720
+    },
+    {
+      "epoch": 254.37908496732027,
+      "grad_norm": 0.29076987504959106,
+      "learning_rate": 9.62144588347823e-05,
+      "loss": 0.1119,
+      "step": 9730
+    },
+    {
+      "epoch": 254.640522875817,
+      "grad_norm": 0.28226718306541443,
+      "learning_rate": 9.604923561931337e-05,
+      "loss": 0.1151,
+      "step": 9740
+    },
+    {
+      "epoch": 254.90196078431373,
+      "grad_norm": 0.3857768177986145,
+      "learning_rate": 9.588402320508342e-05,
+      "loss": 0.1177,
+      "step": 9750
+    },
+    {
+      "epoch": 255.16339869281046,
+      "grad_norm": 0.2802301347255707,
+      "learning_rate": 9.571882204377687e-05,
+      "loss": 0.1194,
+      "step": 9760
+    },
+    {
+      "epoch": 255.4248366013072,
+      "grad_norm": 0.2925792932510376,
+      "learning_rate": 9.555363258704737e-05,
+      "loss": 0.1099,
+      "step": 9770
+    },
+    {
+      "epoch": 255.68627450980392,
+      "grad_norm": 0.3970232605934143,
+      "learning_rate": 9.538845528651665e-05,
+      "loss": 0.1168,
+      "step": 9780
+    },
+    {
+      "epoch": 255.94771241830065,
+      "grad_norm": 0.32095807790756226,
+      "learning_rate": 9.5223290593773e-05,
+      "loss": 0.1202,
+      "step": 9790
+    },
+    {
+      "epoch": 256.2091503267974,
+      "grad_norm": 0.2640714943408966,
+      "learning_rate": 9.50581389603705e-05,
+      "loss": 0.1147,
+      "step": 9800
+    },
+    {
+      "epoch": 256.47058823529414,
+      "grad_norm": 0.31910863518714905,
+      "learning_rate": 9.489300083782737e-05,
+      "loss": 0.1157,
+      "step": 9810
+    },
+    {
+      "epoch": 256.73202614379085,
+      "grad_norm": 0.32858380675315857,
+      "learning_rate": 9.472787667762493e-05,
+      "loss": 0.1166,
+      "step": 9820
+    },
+    {
+      "epoch": 256.9934640522876,
+      "grad_norm": 0.3285580575466156,
+      "learning_rate": 9.456276693120639e-05,
+      "loss": 0.1159,
+      "step": 9830
+    },
+    {
+      "epoch": 257.2549019607843,
+      "grad_norm": 0.35466259717941284,
+      "learning_rate": 9.43976720499754e-05,
+      "loss": 0.1129,
+      "step": 9840
+    },
+    {
+      "epoch": 257.51633986928107,
+      "grad_norm": 0.30083316564559937,
+      "learning_rate": 9.423259248529511e-05,
+      "loss": 0.1155,
+      "step": 9850
+    },
+    {
+      "epoch": 257.77777777777777,
+      "grad_norm": 0.37580639123916626,
+      "learning_rate": 9.406752868848673e-05,
+      "loss": 0.1166,
+      "step": 9860
+    },
+    {
+      "epoch": 258.03921568627453,
+      "grad_norm": 0.36361733078956604,
+      "learning_rate": 9.390248111082842e-05,
+      "loss": 0.1191,
+      "step": 9870
+    },
+    {
+      "epoch": 258.30065359477123,
+      "grad_norm": 0.3550747334957123,
+      "learning_rate": 9.373745020355387e-05,
+      "loss": 0.115,
+      "step": 9880
+    },
+    {
+      "epoch": 258.562091503268,
+      "grad_norm": 0.2830721437931061,
+      "learning_rate": 9.357243641785134e-05,
+      "loss": 0.1162,
+      "step": 9890
+    },
+    {
+      "epoch": 258.8235294117647,
+      "grad_norm": 0.2829933762550354,
+      "learning_rate": 9.340744020486222e-05,
+      "loss": 0.1133,
+      "step": 9900
+    },
+    {
+      "epoch": 259.08496732026146,
+      "grad_norm": 0.29842811822891235,
+      "learning_rate": 9.324246201567984e-05,
+      "loss": 0.1147,
+      "step": 9910
+    },
+    {
+      "epoch": 259.34640522875816,
+      "grad_norm": 0.32608217000961304,
+      "learning_rate": 9.30775023013483e-05,
+      "loss": 0.1143,
+      "step": 9920
+    },
+    {
+      "epoch": 259.6078431372549,
+      "grad_norm": 0.38989517092704773,
+      "learning_rate": 9.291256151286109e-05,
+      "loss": 0.1172,
+      "step": 9930
+    },
+    {
+      "epoch": 259.8692810457516,
+      "grad_norm": 0.3222101330757141,
+      "learning_rate": 9.274764010116008e-05,
+      "loss": 0.1171,
+      "step": 9940
+    },
+    {
+      "epoch": 260.1307189542484,
+      "grad_norm": 0.3502764105796814,
+      "learning_rate": 9.25827385171341e-05,
+      "loss": 0.1135,
+      "step": 9950
+    },
+    {
+      "epoch": 260.3921568627451,
+      "grad_norm": 0.2739408612251282,
+      "learning_rate": 9.241785721161779e-05,
+      "loss": 0.1119,
+      "step": 9960
+    },
+    {
+      "epoch": 260.65359477124184,
+      "grad_norm": 0.289969801902771,
+      "learning_rate": 9.225299663539038e-05,
+      "loss": 0.1165,
+      "step": 9970
+    },
+    {
+      "epoch": 260.91503267973854,
+      "grad_norm": 0.4188239276409149,
+      "learning_rate": 9.20881572391743e-05,
+      "loss": 0.1183,
+      "step": 9980
+    },
+    {
+      "epoch": 261.1764705882353,
+      "grad_norm": 0.30096152424812317,
+      "learning_rate": 9.192333947363423e-05,
+      "loss": 0.1143,
+      "step": 9990
+    },
+    {
+      "epoch": 261.437908496732,
+      "grad_norm": 0.3192864954471588,
+      "learning_rate": 9.175854378937563e-05,
+      "loss": 0.1124,
+      "step": 10000
+    },
+    {
+      "epoch": 261.69934640522877,
+      "grad_norm": 0.3204704523086548,
+      "learning_rate": 9.159377063694365e-05,
+      "loss": 0.1145,
+      "step": 10010
+    },
+    {
+      "epoch": 261.96078431372547,
+      "grad_norm": 0.45152097940444946,
+      "learning_rate": 9.142902046682171e-05,
+      "loss": 0.119,
+      "step": 10020
+    },
+    {
+      "epoch": 262.22222222222223,
+      "grad_norm": 0.3182998597621918,
+      "learning_rate": 9.126429372943053e-05,
+      "loss": 0.1164,
+      "step": 10030
+    },
+    {
+      "epoch": 262.48366013071893,
+      "grad_norm": 0.2740117311477661,
+      "learning_rate": 9.109959087512673e-05,
+      "loss": 0.1142,
+      "step": 10040
+    },
+    {
+      "epoch": 262.7450980392157,
+      "grad_norm": 0.3511454164981842,
+      "learning_rate": 9.093491235420162e-05,
+      "loss": 0.1154,
+      "step": 10050
+    },
+    {
+      "epoch": 263.0065359477124,
+      "grad_norm": 0.3681274652481079,
+      "learning_rate": 9.077025861688003e-05,
+      "loss": 0.1172,
+      "step": 10060
+    },
+    {
+      "epoch": 263.26797385620915,
+      "grad_norm": 0.31222420930862427,
+      "learning_rate": 9.060563011331888e-05,
+      "loss": 0.1122,
+      "step": 10070
+    },
+    {
+      "epoch": 263.52941176470586,
+      "grad_norm": 0.32507622241973877,
+      "learning_rate": 9.044102729360627e-05,
+      "loss": 0.1127,
+      "step": 10080
+    },
+    {
+      "epoch": 263.7908496732026,
+      "grad_norm": 0.4463390111923218,
+      "learning_rate": 9.027645060776006e-05,
+      "loss": 0.1199,
+      "step": 10090
+    },
+    {
+      "epoch": 264.0522875816994,
+      "grad_norm": 0.27414682507514954,
+      "learning_rate": 9.011190050572659e-05,
+      "loss": 0.1166,
+      "step": 10100
+    },
+    {
+      "epoch": 264.3137254901961,
+      "grad_norm": 0.35508808493614197,
+      "learning_rate": 8.994737743737961e-05,
+      "loss": 0.116,
+      "step": 10110
+    },
+    {
+      "epoch": 264.57516339869284,
+      "grad_norm": 0.29134297370910645,
+      "learning_rate": 8.978288185251881e-05,
+      "loss": 0.1134,
+      "step": 10120
+    },
+    {
+      "epoch": 264.83660130718954,
+      "grad_norm": 0.3992556631565094,
+      "learning_rate": 8.961841420086886e-05,
+      "loss": 0.1177,
+      "step": 10130
+    },
+    {
+      "epoch": 265.0980392156863,
+      "grad_norm": 0.24055442214012146,
+      "learning_rate": 8.945397493207809e-05,
+      "loss": 0.1124,
+      "step": 10140
+    },
+    {
+      "epoch": 265.359477124183,
+      "grad_norm": 0.3088826835155487,
+      "learning_rate": 8.92895644957172e-05,
+      "loss": 0.1138,
+      "step": 10150
+    },
+    {
+      "epoch": 265.62091503267976,
+      "grad_norm": 0.4448016285896301,
+      "learning_rate": 8.912518334127795e-05,
+      "loss": 0.1183,
+      "step": 10160
+    },
+    {
+      "epoch": 265.88235294117646,
+      "grad_norm": 0.31662651896476746,
+      "learning_rate": 8.896083191817221e-05,
+      "loss": 0.1166,
+      "step": 10170
+    },
+    {
+      "epoch": 266.1437908496732,
+      "grad_norm": 0.26434990763664246,
+      "learning_rate": 8.879651067573044e-05,
+      "loss": 0.1122,
+      "step": 10180
+    },
+    {
+      "epoch": 266.4052287581699,
+      "grad_norm": 0.25964638590812683,
+      "learning_rate": 8.86322200632007e-05,
+      "loss": 0.1123,
+      "step": 10190
+    },
+    {
+      "epoch": 266.6666666666667,
+      "grad_norm": 0.35823872685432434,
+      "learning_rate": 8.846796052974727e-05,
+      "loss": 0.1156,
+      "step": 10200
+    },
+    {
+      "epoch": 266.9281045751634,
+      "grad_norm": 0.37653812766075134,
+      "learning_rate": 8.830373252444937e-05,
+      "loss": 0.119,
+      "step": 10210
+    },
+    {
+      "epoch": 267.18954248366015,
+      "grad_norm": 0.33142727613449097,
+      "learning_rate": 8.813953649630012e-05,
+      "loss": 0.1158,
+      "step": 10220
+    },
+    {
+      "epoch": 267.45098039215685,
+      "grad_norm": 0.3292408883571625,
+      "learning_rate": 8.797537289420519e-05,
+      "loss": 0.1137,
+      "step": 10230
+    },
+    {
+      "epoch": 267.7124183006536,
+      "grad_norm": 0.40815725922584534,
+      "learning_rate": 8.781124216698161e-05,
+      "loss": 0.1146,
+      "step": 10240
+    },
+    {
+      "epoch": 267.9738562091503,
+      "grad_norm": 0.36364370584487915,
+      "learning_rate": 8.764714476335657e-05,
+      "loss": 0.1189,
+      "step": 10250
+    },
+    {
+      "epoch": 268.2352941176471,
+      "grad_norm": 0.3509305715560913,
+      "learning_rate": 8.748308113196602e-05,
+      "loss": 0.1144,
+      "step": 10260
+    },
+    {
+      "epoch": 268.4967320261438,
+      "grad_norm": 0.31511905789375305,
+      "learning_rate": 8.731905172135369e-05,
+      "loss": 0.113,
+      "step": 10270
+    },
+    {
+      "epoch": 268.75816993464053,
+      "grad_norm": 0.37249597907066345,
+      "learning_rate": 8.715505697996971e-05,
+      "loss": 0.1169,
+      "step": 10280
+    },
+    {
+      "epoch": 269.01960784313724,
+      "grad_norm": 0.3088653087615967,
+      "learning_rate": 8.699109735616952e-05,
+      "loss": 0.1186,
+      "step": 10290
+    },
+    {
+      "epoch": 269.281045751634,
+      "grad_norm": 0.3084110617637634,
+      "learning_rate": 8.68271732982124e-05,
+      "loss": 0.1107,
+      "step": 10300
+    },
+    {
+      "epoch": 269.5424836601307,
+      "grad_norm": 0.31472697854042053,
+      "learning_rate": 8.666328525426045e-05,
+      "loss": 0.117,
+      "step": 10310
+    },
+    {
+      "epoch": 269.80392156862746,
+      "grad_norm": 0.28198713064193726,
+      "learning_rate": 8.649943367237736e-05,
+      "loss": 0.1148,
+      "step": 10320
+    },
+    {
+      "epoch": 270.06535947712416,
+      "grad_norm": 0.3456663191318512,
+      "learning_rate": 8.633561900052708e-05,
+      "loss": 0.1184,
+      "step": 10330
+    },
+    {
+      "epoch": 270.3267973856209,
+      "grad_norm": 0.2773241102695465,
+      "learning_rate": 8.617184168657275e-05,
+      "loss": 0.1117,
+      "step": 10340
+    },
+    {
+      "epoch": 270.5882352941176,
+      "grad_norm": 0.34341636300086975,
+      "learning_rate": 8.600810217827515e-05,
+      "loss": 0.114,
+      "step": 10350
+    },
+    {
+      "epoch": 270.8496732026144,
+      "grad_norm": 0.3856281042098999,
+      "learning_rate": 8.584440092329193e-05,
+      "loss": 0.1174,
+      "step": 10360
+    },
+    {
+      "epoch": 271.1111111111111,
+      "grad_norm": 0.28968098759651184,
+      "learning_rate": 8.568073836917607e-05,
+      "loss": 0.1161,
+      "step": 10370
+    },
+    {
+      "epoch": 271.37254901960785,
+      "grad_norm": 0.3231571912765503,
+      "learning_rate": 8.551711496337469e-05,
+      "loss": 0.1162,
+      "step": 10380
+    },
+    {
+      "epoch": 271.63398692810455,
+      "grad_norm": 0.31247514486312866,
+      "learning_rate": 8.535353115322806e-05,
+      "loss": 0.1143,
+      "step": 10390
+    },
+    {
+      "epoch": 271.8954248366013,
+      "grad_norm": 0.3014558255672455,
+      "learning_rate": 8.518998738596791e-05,
+      "loss": 0.1173,
+      "step": 10400
+    },
+    {
+      "epoch": 272.15686274509807,
+      "grad_norm": 0.27148059010505676,
+      "learning_rate": 8.502648410871675e-05,
+      "loss": 0.11,
+      "step": 10410
+    },
+    {
+      "epoch": 272.41830065359477,
+      "grad_norm": 0.35765767097473145,
+      "learning_rate": 8.486302176848624e-05,
+      "loss": 0.1151,
+      "step": 10420
+    },
+    {
+      "epoch": 272.67973856209153,
+      "grad_norm": 0.3530411124229431,
+      "learning_rate": 8.469960081217627e-05,
+      "loss": 0.1165,
+      "step": 10430
+    },
+    {
+      "epoch": 272.94117647058823,
+      "grad_norm": 0.3553902804851532,
+      "learning_rate": 8.45362216865734e-05,
+      "loss": 0.1165,
+      "step": 10440
+    },
+    {
+      "epoch": 273.202614379085,
+      "grad_norm": 0.2596057653427124,
+      "learning_rate": 8.437288483834997e-05,
+      "loss": 0.1139,
+      "step": 10450
+    },
+    {
+      "epoch": 273.4640522875817,
+      "grad_norm": 0.29487183690071106,
+      "learning_rate": 8.420959071406266e-05,
+      "loss": 0.118,
+      "step": 10460
+    },
+    {
+      "epoch": 273.72549019607845,
+      "grad_norm": 0.3081933557987213,
+      "learning_rate": 8.404633976015134e-05,
+      "loss": 0.1147,
+      "step": 10470
+    },
+    {
+      "epoch": 273.98692810457516,
+      "grad_norm": 0.3125523030757904,
+      "learning_rate": 8.388313242293802e-05,
+      "loss": 0.1136,
+      "step": 10480
+    },
+    {
+      "epoch": 274.2483660130719,
+      "grad_norm": 0.35245898365974426,
+      "learning_rate": 8.371996914862519e-05,
+      "loss": 0.1136,
+      "step": 10490
+    },
+    {
+      "epoch": 274.5098039215686,
+      "grad_norm": 0.36222124099731445,
+      "learning_rate": 8.355685038329504e-05,
+      "loss": 0.1148,
+      "step": 10500
+    },
+    {
+      "epoch": 274.7712418300654,
+      "grad_norm": 0.34575581550598145,
+      "learning_rate": 8.339377657290808e-05,
+      "loss": 0.114,
+      "step": 10510
+    },
+    {
+      "epoch": 275.0326797385621,
+      "grad_norm": 0.4005025625228882,
+      "learning_rate": 8.323074816330183e-05,
+      "loss": 0.1188,
+      "step": 10520
+    },
+    {
+      "epoch": 275.29411764705884,
+      "grad_norm": 0.32579559087753296,
+      "learning_rate": 8.306776560018985e-05,
+      "loss": 0.1138,
+      "step": 10530
+    },
+    {
+      "epoch": 275.55555555555554,
+      "grad_norm": 0.27991876006126404,
+      "learning_rate": 8.29048293291601e-05,
+      "loss": 0.1125,
+      "step": 10540
+    },
+    {
+      "epoch": 275.8169934640523,
+      "grad_norm": 0.3079290986061096,
+      "learning_rate": 8.27419397956742e-05,
+      "loss": 0.1142,
+      "step": 10550
+    },
+    {
+      "epoch": 276.078431372549,
+      "grad_norm": 0.25126445293426514,
+      "learning_rate": 8.257909744506589e-05,
+      "loss": 0.1178,
+      "step": 10560
+    },
+    {
+      "epoch": 276.33986928104576,
+      "grad_norm": 0.2905788719654083,
+      "learning_rate": 8.241630272253998e-05,
+      "loss": 0.1125,
+      "step": 10570
+    },
+    {
+      "epoch": 276.60130718954247,
+      "grad_norm": 0.38739460706710815,
+      "learning_rate": 8.225355607317096e-05,
+      "loss": 0.1152,
+      "step": 10580
+    },
+    {
+      "epoch": 276.8627450980392,
+      "grad_norm": 0.33940622210502625,
+      "learning_rate": 8.2090857941902e-05,
+      "loss": 0.1168,
+      "step": 10590
+    },
+    {
+      "epoch": 277.12418300653593,
+      "grad_norm": 0.2534843981266022,
+      "learning_rate": 8.192820877354357e-05,
+      "loss": 0.1112,
+      "step": 10600
+    },
+    {
+      "epoch": 277.3856209150327,
+      "grad_norm": 0.3085792660713196,
+      "learning_rate": 8.176560901277229e-05,
+      "loss": 0.1165,
+      "step": 10610
+    },
+    {
+      "epoch": 277.6470588235294,
+      "grad_norm": 0.32212698459625244,
+      "learning_rate": 8.16030591041297e-05,
+      "loss": 0.1138,
+      "step": 10620
+    },
+    {
+      "epoch": 277.90849673202615,
+      "grad_norm": 0.31586194038391113,
+      "learning_rate": 8.144055949202101e-05,
+      "loss": 0.1186,
+      "step": 10630
+    },
+    {
+      "epoch": 278.16993464052285,
+      "grad_norm": 0.2730450928211212,
+      "learning_rate": 8.127811062071398e-05,
+      "loss": 0.1163,
+      "step": 10640
+    },
+    {
+      "epoch": 278.4313725490196,
+      "grad_norm": 0.2704620659351349,
+      "learning_rate": 8.111571293433764e-05,
+      "loss": 0.1104,
+      "step": 10650
+    },
+    {
+      "epoch": 278.6928104575163,
+      "grad_norm": 0.3569394052028656,
+      "learning_rate": 8.095336687688102e-05,
+      "loss": 0.1171,
+      "step": 10660
+    },
+    {
+      "epoch": 278.9542483660131,
+      "grad_norm": 0.28940895199775696,
+      "learning_rate": 8.079107289219209e-05,
+      "loss": 0.1172,
+      "step": 10670
+    },
+    {
+      "epoch": 279.2156862745098,
+      "grad_norm": 0.30181649327278137,
+      "learning_rate": 8.062883142397635e-05,
+      "loss": 0.1124,
+      "step": 10680
+    },
+    {
+      "epoch": 279.47712418300654,
+      "grad_norm": 0.3073585629463196,
+      "learning_rate": 8.046664291579584e-05,
+      "loss": 0.1137,
+      "step": 10690
+    },
+    {
+      "epoch": 279.73856209150324,
+      "grad_norm": 0.30750972032546997,
+      "learning_rate": 8.03045078110677e-05,
+      "loss": 0.115,
+      "step": 10700
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 0.33145976066589355,
+      "learning_rate": 8.014242655306315e-05,
+      "loss": 0.1187,
+      "step": 10710
+    },
+    {
+      "epoch": 280.26143790849676,
+      "grad_norm": 0.2784498631954193,
+      "learning_rate": 7.998039958490613e-05,
+      "loss": 0.1113,
+      "step": 10720
+    },
+    {
+      "epoch": 280.52287581699346,
+      "grad_norm": 0.3815988302230835,
+      "learning_rate": 7.981842734957221e-05,
+      "loss": 0.1154,
+      "step": 10730
+    },
+    {
+      "epoch": 280.7843137254902,
+      "grad_norm": 0.35885512828826904,
+      "learning_rate": 7.965651028988726e-05,
+      "loss": 0.1192,
+      "step": 10740
+    },
+    {
+      "epoch": 281.0457516339869,
+      "grad_norm": 0.3109668791294098,
+      "learning_rate": 7.949464884852638e-05,
+      "loss": 0.114,
+      "step": 10750
+    },
+    {
+      "epoch": 281.3071895424837,
+      "grad_norm": 0.2830188274383545,
+      "learning_rate": 7.933284346801258e-05,
+      "loss": 0.1111,
+      "step": 10760
+    },
+    {
+      "epoch": 281.5686274509804,
+      "grad_norm": 0.30252325534820557,
+      "learning_rate": 7.917109459071553e-05,
+      "loss": 0.113,
+      "step": 10770
+    },
+    {
+      "epoch": 281.83006535947715,
+      "grad_norm": 0.3681179881095886,
+      "learning_rate": 7.900940265885052e-05,
+      "loss": 0.1172,
+      "step": 10780
+    },
+    {
+      "epoch": 282.09150326797385,
+      "grad_norm": 0.3035522401332855,
+      "learning_rate": 7.884776811447712e-05,
+      "loss": 0.1148,
+      "step": 10790
+    },
+    {
+      "epoch": 282.3529411764706,
+      "grad_norm": 0.2895407974720001,
+      "learning_rate": 7.8686191399498e-05,
+      "loss": 0.1125,
+      "step": 10800
+    },
+    {
+      "epoch": 282.6143790849673,
+      "grad_norm": 0.37216058373451233,
+      "learning_rate": 7.852467295565775e-05,
+      "loss": 0.1155,
+      "step": 10810
+    },
+    {
+      "epoch": 282.87581699346407,
+      "grad_norm": 0.3305184543132782,
+      "learning_rate": 7.836321322454159e-05,
+      "loss": 0.1163,
+      "step": 10820
+    },
+    {
+      "epoch": 283.1372549019608,
+      "grad_norm": 0.30414262413978577,
+      "learning_rate": 7.820181264757427e-05,
+      "loss": 0.1154,
+      "step": 10830
+    },
+    {
+      "epoch": 283.39869281045753,
+      "grad_norm": 0.2971784174442291,
+      "learning_rate": 7.80404716660188e-05,
+      "loss": 0.1126,
+      "step": 10840
+    },
+    {
+      "epoch": 283.66013071895424,
+      "grad_norm": 0.36141228675842285,
+      "learning_rate": 7.787919072097531e-05,
+      "loss": 0.1155,
+      "step": 10850
+    },
+    {
+      "epoch": 283.921568627451,
+      "grad_norm": 0.2840782403945923,
+      "learning_rate": 7.771797025337968e-05,
+      "loss": 0.1198,
+      "step": 10860
+    },
+    {
+      "epoch": 284.1830065359477,
+      "grad_norm": 0.2825259268283844,
+      "learning_rate": 7.755681070400253e-05,
+      "loss": 0.1123,
+      "step": 10870
+    },
+    {
+      "epoch": 284.44444444444446,
+      "grad_norm": 0.3692414462566376,
+      "learning_rate": 7.739571251344794e-05,
+      "loss": 0.1157,
+      "step": 10880
+    },
+    {
+      "epoch": 284.70588235294116,
+      "grad_norm": 0.2816650867462158,
+      "learning_rate": 7.723467612215219e-05,
+      "loss": 0.1159,
+      "step": 10890
+    },
+    {
+      "epoch": 284.9673202614379,
+      "grad_norm": 0.33248651027679443,
+      "learning_rate": 7.707370197038265e-05,
+      "loss": 0.1158,
+      "step": 10900
+    },
+    {
+      "epoch": 285.2287581699346,
+      "grad_norm": 0.3684132695198059,
+      "learning_rate": 7.691279049823646e-05,
+      "loss": 0.1138,
+      "step": 10910
+    },
+    {
+      "epoch": 285.4901960784314,
+      "grad_norm": 0.28321191668510437,
+      "learning_rate": 7.675194214563948e-05,
+      "loss": 0.1137,
+      "step": 10920
+    },
+    {
+      "epoch": 285.7516339869281,
+      "grad_norm": 0.39478781819343567,
+      "learning_rate": 7.659115735234494e-05,
+      "loss": 0.1193,
+      "step": 10930
+    },
+    {
+      "epoch": 286.01307189542484,
+      "grad_norm": 0.3251636028289795,
+      "learning_rate": 7.643043655793235e-05,
+      "loss": 0.1118,
+      "step": 10940
+    },
+    {
+      "epoch": 286.27450980392155,
+      "grad_norm": 0.32104846835136414,
+      "learning_rate": 7.62697802018062e-05,
+      "loss": 0.1134,
+      "step": 10950
+    },
+    {
+      "epoch": 286.5359477124183,
+      "grad_norm": 0.3238525092601776,
+      "learning_rate": 7.610918872319483e-05,
+      "loss": 0.1126,
+      "step": 10960
+    },
+    {
+      "epoch": 286.797385620915,
+      "grad_norm": 0.3236340284347534,
+      "learning_rate": 7.594866256114921e-05,
+      "loss": 0.1178,
+      "step": 10970
+    },
+    {
+      "epoch": 287.05882352941177,
+      "grad_norm": 0.2717225253582001,
+      "learning_rate": 7.578820215454178e-05,
+      "loss": 0.1153,
+      "step": 10980
+    },
+    {
+      "epoch": 287.32026143790847,
+      "grad_norm": 0.34041455388069153,
+      "learning_rate": 7.562780794206514e-05,
+      "loss": 0.1129,
+      "step": 10990
+    },
+    {
+      "epoch": 287.58169934640523,
+      "grad_norm": 0.27632108330726624,
+      "learning_rate": 7.546748036223091e-05,
+      "loss": 0.1154,
+      "step": 11000
+    },
+    {
+      "epoch": 287.84313725490193,
+      "grad_norm": 0.4436096251010895,
+      "learning_rate": 7.530721985336861e-05,
+      "loss": 0.1157,
+      "step": 11010
+    },
+    {
+      "epoch": 288.1045751633987,
+      "grad_norm": 0.3542175889015198,
+      "learning_rate": 7.514702685362434e-05,
+      "loss": 0.1161,
+      "step": 11020
+    },
+    {
+      "epoch": 288.36601307189545,
+      "grad_norm": 0.32913169264793396,
+      "learning_rate": 7.498690180095963e-05,
+      "loss": 0.115,
+      "step": 11030
+    },
+    {
+      "epoch": 288.62745098039215,
+      "grad_norm": 0.4230281710624695,
+      "learning_rate": 7.48268451331503e-05,
+      "loss": 0.1147,
+      "step": 11040
+    },
+    {
+      "epoch": 288.8888888888889,
+      "grad_norm": 0.31047868728637695,
+      "learning_rate": 7.466685728778513e-05,
+      "loss": 0.1136,
+      "step": 11050
+    },
+    {
+      "epoch": 289.1503267973856,
+      "grad_norm": 0.2447769194841385,
+      "learning_rate": 7.450693870226478e-05,
+      "loss": 0.1124,
+      "step": 11060
+    },
+    {
+      "epoch": 289.4117647058824,
+      "grad_norm": 0.2783534824848175,
+      "learning_rate": 7.434708981380057e-05,
+      "loss": 0.1145,
+      "step": 11070
+    },
+    {
+      "epoch": 289.6732026143791,
+      "grad_norm": 0.2729416787624359,
+      "learning_rate": 7.418731105941328e-05,
+      "loss": 0.1152,
+      "step": 11080
+    },
+    {
+      "epoch": 289.93464052287584,
+      "grad_norm": 0.29936328530311584,
+      "learning_rate": 7.402760287593189e-05,
+      "loss": 0.1175,
+      "step": 11090
+    },
+    {
+      "epoch": 290.19607843137254,
+      "grad_norm": 0.3042921721935272,
+      "learning_rate": 7.386796569999246e-05,
+      "loss": 0.1128,
+      "step": 11100
+    },
+    {
+      "epoch": 290.4575163398693,
+      "grad_norm": 0.3261522948741913,
+      "learning_rate": 7.370839996803697e-05,
+      "loss": 0.1144,
+      "step": 11110
+    },
+    {
+      "epoch": 290.718954248366,
+      "grad_norm": 0.30387112498283386,
+      "learning_rate": 7.354890611631202e-05,
+      "loss": 0.1139,
+      "step": 11120
+    },
+    {
+      "epoch": 290.98039215686276,
+      "grad_norm": 0.3516126275062561,
+      "learning_rate": 7.338948458086774e-05,
+      "loss": 0.1167,
+      "step": 11130
+    },
+    {
+      "epoch": 291.24183006535947,
+      "grad_norm": 0.2861202359199524,
+      "learning_rate": 7.323013579755647e-05,
+      "loss": 0.1138,
+      "step": 11140
+    },
+    {
+      "epoch": 291.5032679738562,
+      "grad_norm": 0.3707253634929657,
+      "learning_rate": 7.307086020203173e-05,
+      "loss": 0.1136,
+      "step": 11150
+    },
+    {
+      "epoch": 291.7647058823529,
+      "grad_norm": 0.34529221057891846,
+      "learning_rate": 7.29116582297469e-05,
+      "loss": 0.1167,
+      "step": 11160
+    },
+    {
+      "epoch": 292.0261437908497,
+      "grad_norm": 0.2350694239139557,
+      "learning_rate": 7.275253031595413e-05,
+      "loss": 0.1139,
+      "step": 11170
+    },
+    {
+      "epoch": 292.2875816993464,
+      "grad_norm": 0.35073357820510864,
+      "learning_rate": 7.259347689570304e-05,
+      "loss": 0.1141,
+      "step": 11180
+    },
+    {
+      "epoch": 292.54901960784315,
+      "grad_norm": 0.3456834554672241,
+      "learning_rate": 7.243449840383958e-05,
+      "loss": 0.1117,
+      "step": 11190
+    },
+    {
+      "epoch": 292.81045751633985,
+      "grad_norm": 0.3201320171356201,
+      "learning_rate": 7.227559527500489e-05,
+      "loss": 0.1177,
+      "step": 11200
+    },
+    {
+      "epoch": 293.0718954248366,
+      "grad_norm": 0.2568330466747284,
+      "learning_rate": 7.211676794363407e-05,
+      "loss": 0.1144,
+      "step": 11210
+    },
+    {
+      "epoch": 293.3333333333333,
+      "grad_norm": 0.29000136256217957,
+      "learning_rate": 7.1958016843955e-05,
+      "loss": 0.1121,
+      "step": 11220
+    },
+    {
+      "epoch": 293.5947712418301,
+      "grad_norm": 0.30854013562202454,
+      "learning_rate": 7.179934240998706e-05,
+      "loss": 0.1206,
+      "step": 11230
+    },
+    {
+      "epoch": 293.8562091503268,
+      "grad_norm": 0.27561551332473755,
+      "learning_rate": 7.164074507554015e-05,
+      "loss": 0.1137,
+      "step": 11240
+    },
+    {
+      "epoch": 294.11764705882354,
+      "grad_norm": 0.28580352663993835,
+      "learning_rate": 7.148222527421331e-05,
+      "loss": 0.1125,
+      "step": 11250
+    },
+    {
+      "epoch": 294.37908496732024,
+      "grad_norm": 0.29733917117118835,
+      "learning_rate": 7.132378343939361e-05,
+      "loss": 0.1142,
+      "step": 11260
+    },
+    {
+      "epoch": 294.640522875817,
+      "grad_norm": 0.3131226897239685,
+      "learning_rate": 7.116542000425501e-05,
+      "loss": 0.1151,
+      "step": 11270
+    },
+    {
+      "epoch": 294.9019607843137,
+      "grad_norm": 0.2978150248527527,
+      "learning_rate": 7.100713540175706e-05,
+      "loss": 0.1152,
+      "step": 11280
+    },
+    {
+      "epoch": 295.16339869281046,
+      "grad_norm": 0.31860676407814026,
+      "learning_rate": 7.084893006464383e-05,
+      "loss": 0.1136,
+      "step": 11290
+    },
+    {
+      "epoch": 295.42483660130716,
+      "grad_norm": 0.36646994948387146,
+      "learning_rate": 7.069080442544267e-05,
+      "loss": 0.1127,
+      "step": 11300
+    },
+    {
+      "epoch": 295.6862745098039,
+      "grad_norm": 0.3542885184288025,
+      "learning_rate": 7.053275891646303e-05,
+      "loss": 0.1134,
+      "step": 11310
+    },
+    {
+      "epoch": 295.9477124183006,
+      "grad_norm": 0.46107035875320435,
+      "learning_rate": 7.037479396979535e-05,
+      "loss": 0.1177,
+      "step": 11320
+    },
+    {
+      "epoch": 296.2091503267974,
+      "grad_norm": 0.32113420963287354,
+      "learning_rate": 7.02169100173097e-05,
+      "loss": 0.1149,
+      "step": 11330
+    },
+    {
+      "epoch": 296.47058823529414,
+      "grad_norm": 0.3642306625843048,
+      "learning_rate": 7.005910749065478e-05,
+      "loss": 0.1157,
+      "step": 11340
+    },
+    {
+      "epoch": 296.73202614379085,
+      "grad_norm": 0.35815343260765076,
+      "learning_rate": 6.990138682125671e-05,
+      "loss": 0.1135,
+      "step": 11350
+    },
+    {
+      "epoch": 296.9934640522876,
+      "grad_norm": 0.365302175283432,
+      "learning_rate": 6.974374844031779e-05,
+      "loss": 0.1158,
+      "step": 11360
+    },
+    {
+      "epoch": 297.2549019607843,
+      "grad_norm": 0.26450252532958984,
+      "learning_rate": 6.958619277881524e-05,
+      "loss": 0.1148,
+      "step": 11370
+    },
+    {
+      "epoch": 297.51633986928107,
+      "grad_norm": 0.29163211584091187,
+      "learning_rate": 6.942872026750029e-05,
+      "loss": 0.1134,
+      "step": 11380
+    },
+    {
+      "epoch": 297.77777777777777,
+      "grad_norm": 0.3450673520565033,
+      "learning_rate": 6.927133133689678e-05,
+      "loss": 0.1158,
+      "step": 11390
+    },
+    {
+      "epoch": 298.03921568627453,
+      "grad_norm": 0.2773265838623047,
+      "learning_rate": 6.911402641730003e-05,
+      "loss": 0.1136,
+      "step": 11400
+    },
+    {
+      "epoch": 298.30065359477123,
+      "grad_norm": 0.29598814249038696,
+      "learning_rate": 6.895680593877571e-05,
+      "loss": 0.1168,
+      "step": 11410
+    },
+    {
+      "epoch": 298.562091503268,
+      "grad_norm": 0.2809271812438965,
+      "learning_rate": 6.879967033115853e-05,
+      "loss": 0.1142,
+      "step": 11420
+    },
+    {
+      "epoch": 298.8235294117647,
+      "grad_norm": 0.2930368185043335,
+      "learning_rate": 6.864262002405129e-05,
+      "loss": 0.1148,
+      "step": 11430
+    },
+    {
+      "epoch": 299.08496732026146,
+      "grad_norm": 0.25596320629119873,
+      "learning_rate": 6.848565544682352e-05,
+      "loss": 0.1114,
+      "step": 11440
+    },
+    {
+      "epoch": 299.34640522875816,
+      "grad_norm": 0.26733312010765076,
+      "learning_rate": 6.832877702861037e-05,
+      "loss": 0.1138,
+      "step": 11450
+    },
+    {
+      "epoch": 299.6078431372549,
+      "grad_norm": 0.33392226696014404,
+      "learning_rate": 6.817198519831154e-05,
+      "loss": 0.116,
+      "step": 11460
+    },
+    {
+      "epoch": 299.8692810457516,
+      "grad_norm": 0.28333500027656555,
+      "learning_rate": 6.801528038458974e-05,
+      "loss": 0.1129,
+      "step": 11470
+    },
+    {
+      "epoch": 300.1307189542484,
+      "grad_norm": 0.28260430693626404,
+      "learning_rate": 6.785866301587007e-05,
+      "loss": 0.1151,
+      "step": 11480
+    },
+    {
+      "epoch": 300.3921568627451,
+      "grad_norm": 0.35571715235710144,
+      "learning_rate": 6.770213352033839e-05,
+      "loss": 0.1129,
+      "step": 11490
+    },
+    {
+      "epoch": 300.65359477124184,
+      "grad_norm": 0.3103904724121094,
+      "learning_rate": 6.754569232594042e-05,
+      "loss": 0.1145,
+      "step": 11500
+    },
+    {
+      "epoch": 300.91503267973854,
+      "grad_norm": 0.3481941223144531,
+      "learning_rate": 6.738933986038033e-05,
+      "loss": 0.1163,
+      "step": 11510
+    },
+    {
+      "epoch": 301.1764705882353,
+      "grad_norm": 0.2778976559638977,
+      "learning_rate": 6.72330765511198e-05,
+      "loss": 0.1141,
+      "step": 11520
+    },
+    {
+      "epoch": 301.437908496732,
+      "grad_norm": 0.472317099571228,
+      "learning_rate": 6.70769028253768e-05,
+      "loss": 0.1159,
+      "step": 11530
+    },
+    {
+      "epoch": 301.69934640522877,
+      "grad_norm": 0.3168316185474396,
+      "learning_rate": 6.692081911012431e-05,
+      "loss": 0.117,
+      "step": 11540
+    },
+    {
+      "epoch": 301.96078431372547,
+      "grad_norm": 0.2718193531036377,
+      "learning_rate": 6.676482583208929e-05,
+      "loss": 0.114,
+      "step": 11550
+    },
+    {
+      "epoch": 302.22222222222223,
+      "grad_norm": 0.32255101203918457,
+      "learning_rate": 6.660892341775132e-05,
+      "loss": 0.1129,
+      "step": 11560
+    },
+    {
+      "epoch": 302.48366013071893,
+      "grad_norm": 0.4141261577606201,
+      "learning_rate": 6.645311229334167e-05,
+      "loss": 0.1154,
+      "step": 11570
+    },
+    {
+      "epoch": 302.7450980392157,
+      "grad_norm": 0.28321996331214905,
+      "learning_rate": 6.629739288484204e-05,
+      "loss": 0.1125,
+      "step": 11580
+    },
+    {
+      "epoch": 303.0065359477124,
+      "grad_norm": 0.27643659710884094,
+      "learning_rate": 6.614176561798335e-05,
+      "loss": 0.1153,
+      "step": 11590
+    },
+    {
+      "epoch": 303.26797385620915,
+      "grad_norm": 0.3224067986011505,
+      "learning_rate": 6.598623091824461e-05,
+      "loss": 0.1103,
+      "step": 11600
+    },
+    {
+      "epoch": 303.52941176470586,
+      "grad_norm": 0.31204745173454285,
+      "learning_rate": 6.583078921085167e-05,
+      "loss": 0.114,
+      "step": 11610
+    },
+    {
+      "epoch": 303.7908496732026,
+      "grad_norm": 0.3363180458545685,
+      "learning_rate": 6.567544092077631e-05,
+      "loss": 0.1164,
+      "step": 11620
+    },
+    {
+      "epoch": 304.0522875816994,
+      "grad_norm": 0.3073705732822418,
+      "learning_rate": 6.55357076825483e-05,
+      "loss": 0.1173,
+      "step": 11630
+    },
+    {
+      "epoch": 304.3137254901961,
+      "grad_norm": 0.38721707463264465,
+      "learning_rate": 6.538053805525763e-05,
+      "loss": 0.1128,
+      "step": 11640
+    },
+    {
+      "epoch": 304.57516339869284,
+      "grad_norm": 0.3045378625392914,
+      "learning_rate": 6.522546307625399e-05,
+      "loss": 0.1154,
+      "step": 11650
+    },
+    {
+      "epoch": 304.83660130718954,
+      "grad_norm": 0.29825395345687866,
+      "learning_rate": 6.507048316950648e-05,
+      "loss": 0.1135,
+      "step": 11660
+    },
+    {
+      "epoch": 305.0980392156863,
+      "grad_norm": 0.32951053977012634,
+      "learning_rate": 6.491559875872415e-05,
+      "loss": 0.1168,
+      "step": 11670
+    },
+    {
+      "epoch": 305.359477124183,
+      "grad_norm": 0.25783228874206543,
+      "learning_rate": 6.476081026735513e-05,
+      "loss": 0.1094,
+      "step": 11680
+    },
+    {
+      "epoch": 305.62091503267976,
+      "grad_norm": 0.3310535252094269,
+      "learning_rate": 6.460611811858521e-05,
+      "loss": 0.1162,
+      "step": 11690
+    },
+    {
+      "epoch": 305.88235294117646,
+      "grad_norm": 0.3044494390487671,
+      "learning_rate": 6.445152273533687e-05,
+      "loss": 0.1172,
+      "step": 11700
+    },
+    {
+      "epoch": 306.1437908496732,
+      "grad_norm": 0.2858486771583557,
+      "learning_rate": 6.429702454026798e-05,
+      "loss": 0.1128,
+      "step": 11710
+    },
+    {
+      "epoch": 306.4052287581699,
+      "grad_norm": 0.28763970732688904,
+      "learning_rate": 6.414262395577065e-05,
+      "loss": 0.1123,
+      "step": 11720
+    },
+    {
+      "epoch": 306.6666666666667,
+      "grad_norm": 0.33545467257499695,
+      "learning_rate": 6.398832140397022e-05,
+      "loss": 0.117,
+      "step": 11730
+    },
+    {
+      "epoch": 306.9281045751634,
+      "grad_norm": 0.33642587065696716,
+      "learning_rate": 6.383411730672394e-05,
+      "loss": 0.1168,
+      "step": 11740
+    },
+    {
+      "epoch": 307.18954248366015,
+      "grad_norm": 0.3563029170036316,
+      "learning_rate": 6.368001208561998e-05,
+      "loss": 0.1135,
+      "step": 11750
+    },
+    {
+      "epoch": 307.45098039215685,
+      "grad_norm": 0.31238290667533875,
+      "learning_rate": 6.352600616197615e-05,
+      "loss": 0.1139,
+      "step": 11760
+    },
+    {
+      "epoch": 307.7124183006536,
+      "grad_norm": 0.3694748282432556,
+      "learning_rate": 6.337209995683867e-05,
+      "loss": 0.1166,
+      "step": 11770
+    },
+    {
+      "epoch": 307.9738562091503,
+      "grad_norm": 0.2764953672885895,
+      "learning_rate": 6.321829389098126e-05,
+      "loss": 0.1135,
+      "step": 11780
+    },
+    {
+      "epoch": 308.2352941176471,
+      "grad_norm": 0.28875818848609924,
+      "learning_rate": 6.306458838490385e-05,
+      "loss": 0.1121,
+      "step": 11790
+    },
+    {
+      "epoch": 308.4967320261438,
+      "grad_norm": 0.3236418068408966,
+      "learning_rate": 6.291098385883146e-05,
+      "loss": 0.1093,
+      "step": 11800
+    },
+    {
+      "epoch": 308.75816993464053,
+      "grad_norm": 0.2844972014427185,
+      "learning_rate": 6.275748073271292e-05,
+      "loss": 0.1184,
+      "step": 11810
+    },
+    {
+      "epoch": 309.01960784313724,
+      "grad_norm": 0.39473089575767517,
+      "learning_rate": 6.260407942621998e-05,
+      "loss": 0.1169,
+      "step": 11820
+    },
+    {
+      "epoch": 309.281045751634,
+      "grad_norm": 0.2659190893173218,
+      "learning_rate": 6.245078035874591e-05,
+      "loss": 0.1132,
+      "step": 11830
+    },
+    {
+      "epoch": 309.5424836601307,
+      "grad_norm": 0.28653818368911743,
+      "learning_rate": 6.22975839494045e-05,
+      "loss": 0.1137,
+      "step": 11840
+    },
+    {
+      "epoch": 309.80392156862746,
+      "grad_norm": 0.311483770608902,
+      "learning_rate": 6.214449061702898e-05,
+      "loss": 0.1127,
+      "step": 11850
+    },
+    {
+      "epoch": 310.06535947712416,
+      "grad_norm": 0.2597990930080414,
+      "learning_rate": 6.199150078017057e-05,
+      "loss": 0.1135,
+      "step": 11860
+    },
+    {
+      "epoch": 310.3267973856209,
+      "grad_norm": 0.32311543822288513,
+      "learning_rate": 6.183861485709765e-05,
+      "loss": 0.1154,
+      "step": 11870
+    },
+    {
+      "epoch": 310.5882352941176,
+      "grad_norm": 0.3250320553779602,
+      "learning_rate": 6.168583326579456e-05,
+      "loss": 0.1135,
+      "step": 11880
+    },
+    {
+      "epoch": 310.8496732026144,
+      "grad_norm": 0.2815916836261749,
+      "learning_rate": 6.153315642396025e-05,
+      "loss": 0.1145,
+      "step": 11890
+    },
+    {
+      "epoch": 311.1111111111111,
+      "grad_norm": 0.3453966975212097,
+      "learning_rate": 6.13805847490075e-05,
+      "loss": 0.1141,
+      "step": 11900
+    },
+    {
+      "epoch": 311.37254901960785,
+      "grad_norm": 0.3138068914413452,
+      "learning_rate": 6.122811865806131e-05,
+      "loss": 0.1127,
+      "step": 11910
+    },
+    {
+      "epoch": 311.63398692810455,
+      "grad_norm": 0.3160851299762726,
+      "learning_rate": 6.107575856795822e-05,
+      "loss": 0.1143,
+      "step": 11920
+    },
+    {
+      "epoch": 311.8954248366013,
+      "grad_norm": 0.3921455144882202,
+      "learning_rate": 6.0923504895244875e-05,
+      "loss": 0.1145,
+      "step": 11930
+    },
+    {
+      "epoch": 312.15686274509807,
+      "grad_norm": 0.31224459409713745,
+      "learning_rate": 6.077135805617705e-05,
+      "loss": 0.1113,
+      "step": 11940
+    },
+    {
+      "epoch": 312.41830065359477,
+      "grad_norm": 0.330119788646698,
+      "learning_rate": 6.061931846671833e-05,
+      "loss": 0.1154,
+      "step": 11950
+    },
+    {
+      "epoch": 312.67973856209153,
+      "grad_norm": 0.3644309639930725,
+      "learning_rate": 6.046738654253918e-05,
+      "loss": 0.1165,
+      "step": 11960
+    },
+    {
+      "epoch": 312.94117647058823,
+      "grad_norm": 0.30337658524513245,
+      "learning_rate": 6.031556269901567e-05,
+      "loss": 0.1154,
+      "step": 11970
+    },
+    {
+      "epoch": 313.202614379085,
+      "grad_norm": 0.31728994846343994,
+      "learning_rate": 6.0163847351228395e-05,
+      "loss": 0.1113,
+      "step": 11980
+    },
+    {
+      "epoch": 313.4640522875817,
+      "grad_norm": 0.32171520590782166,
+      "learning_rate": 6.0012240913961334e-05,
+      "loss": 0.1156,
+      "step": 11990
+    },
+    {
+      "epoch": 313.72549019607845,
+      "grad_norm": 0.2858985364437103,
+      "learning_rate": 5.986074380170068e-05,
+      "loss": 0.1139,
+      "step": 12000
+    },
+    {
+      "epoch": 313.98692810457516,
+      "grad_norm": 0.32188278436660767,
+      "learning_rate": 5.9709356428633746e-05,
+      "loss": 0.1163,
+      "step": 12010
+    },
+    {
+      "epoch": 314.2483660130719,
+      "grad_norm": 0.2941359877586365,
+      "learning_rate": 5.955807920864784e-05,
+      "loss": 0.1123,
+      "step": 12020
+    },
+    {
+      "epoch": 314.5098039215686,
+      "grad_norm": 0.29654598236083984,
+      "learning_rate": 5.940691255532912e-05,
+      "loss": 0.1116,
+      "step": 12030
+    },
+    {
+      "epoch": 314.7712418300654,
+      "grad_norm": 0.29309600591659546,
+      "learning_rate": 5.9255856881961444e-05,
+      "loss": 0.113,
+      "step": 12040
+    },
+    {
+      "epoch": 315.0326797385621,
+      "grad_norm": 0.3037513792514801,
+      "learning_rate": 5.910491260152522e-05,
+      "loss": 0.1196,
+      "step": 12050
+    },
+    {
+      "epoch": 315.29411764705884,
+      "grad_norm": 0.37008363008499146,
+      "learning_rate": 5.8954080126696366e-05,
+      "loss": 0.1123,
+      "step": 12060
+    },
+    {
+      "epoch": 315.55555555555554,
+      "grad_norm": 0.2903066575527191,
+      "learning_rate": 5.880335986984512e-05,
+      "loss": 0.1139,
+      "step": 12070
+    },
+    {
+      "epoch": 315.8169934640523,
+      "grad_norm": 0.42981478571891785,
+      "learning_rate": 5.865275224303491e-05,
+      "loss": 0.1145,
+      "step": 12080
+    },
+    {
+      "epoch": 316.078431372549,
+      "grad_norm": 0.2848221957683563,
+      "learning_rate": 5.850225765802122e-05,
+      "loss": 0.1178,
+      "step": 12090
+    },
+    {
+      "epoch": 316.33986928104576,
+      "grad_norm": 0.3206409811973572,
+      "learning_rate": 5.835187652625047e-05,
+      "loss": 0.1125,
+      "step": 12100
+    },
+    {
+      "epoch": 316.60130718954247,
+      "grad_norm": 0.30614250898361206,
+      "learning_rate": 5.820160925885902e-05,
+      "loss": 0.1139,
+      "step": 12110
+    },
+    {
+      "epoch": 316.8627450980392,
+      "grad_norm": 0.32410693168640137,
+      "learning_rate": 5.8051456266671746e-05,
+      "loss": 0.1144,
+      "step": 12120
+    },
+    {
+      "epoch": 317.12418300653593,
+      "grad_norm": 0.3014141619205475,
+      "learning_rate": 5.790141796020132e-05,
+      "loss": 0.114,
+      "step": 12130
+    },
+    {
+      "epoch": 317.3856209150327,
+      "grad_norm": 0.30295130610466003,
+      "learning_rate": 5.7751494749646575e-05,
+      "loss": 0.115,
+      "step": 12140
+    },
+    {
+      "epoch": 317.6470588235294,
+      "grad_norm": 0.3472057282924652,
+      "learning_rate": 5.7601687044891925e-05,
+      "loss": 0.1126,
+      "step": 12150
+    },
+    {
+      "epoch": 317.90849673202615,
+      "grad_norm": 0.2875531315803528,
+      "learning_rate": 5.745199525550596e-05,
+      "loss": 0.1153,
+      "step": 12160
+    },
+    {
+      "epoch": 318.16993464052285,
+      "grad_norm": 0.3130854070186615,
+      "learning_rate": 5.730241979074025e-05,
+      "loss": 0.1143,
+      "step": 12170
+    },
+    {
+      "epoch": 318.4313725490196,
+      "grad_norm": 0.27834898233413696,
+      "learning_rate": 5.71529610595285e-05,
+      "loss": 0.1129,
+      "step": 12180
+    },
+    {
+      "epoch": 318.6928104575163,
+      "grad_norm": 0.37341535091400146,
+      "learning_rate": 5.7003619470485016e-05,
+      "loss": 0.1133,
+      "step": 12190
+    },
+    {
+      "epoch": 318.9542483660131,
+      "grad_norm": 0.31970056891441345,
+      "learning_rate": 5.6854395431904094e-05,
+      "loss": 0.1151,
+      "step": 12200
+    },
+    {
+      "epoch": 319.2156862745098,
+      "grad_norm": 0.3472343683242798,
+      "learning_rate": 5.6705289351758584e-05,
+      "loss": 0.1109,
+      "step": 12210
+    },
+    {
+      "epoch": 319.47712418300654,
+      "grad_norm": 0.3212395906448364,
+      "learning_rate": 5.6556301637698785e-05,
+      "loss": 0.1119,
+      "step": 12220
+    },
+    {
+      "epoch": 319.73856209150324,
+      "grad_norm": 0.40940025448799133,
+      "learning_rate": 5.6407432697051424e-05,
+      "loss": 0.1174,
+      "step": 12230
+    },
+    {
+      "epoch": 320.0,
+      "grad_norm": 0.3223443925380707,
+      "learning_rate": 5.625868293681844e-05,
+      "loss": 0.1155,
+      "step": 12240
+    },
+    {
+      "epoch": 320.26143790849676,
+      "grad_norm": 0.295582115650177,
+      "learning_rate": 5.611005276367605e-05,
+      "loss": 0.1145,
+      "step": 12250
+    },
+    {
+      "epoch": 320.52287581699346,
+      "grad_norm": 0.3024216592311859,
+      "learning_rate": 5.596154258397353e-05,
+      "loss": 0.1104,
+      "step": 12260
+    },
+    {
+      "epoch": 320.7843137254902,
+      "grad_norm": 0.3048156797885895,
+      "learning_rate": 5.581315280373195e-05,
+      "loss": 0.1134,
+      "step": 12270
+    },
+    {
+      "epoch": 321.0457516339869,
+      "grad_norm": 0.256906121969223,
+      "learning_rate": 5.566488382864334e-05,
+      "loss": 0.1156,
+      "step": 12280
+    },
+    {
+      "epoch": 321.3071895424837,
+      "grad_norm": 0.2839376628398895,
+      "learning_rate": 5.55167360640694e-05,
+      "loss": 0.1141,
+      "step": 12290
+    },
+    {
+      "epoch": 321.5686274509804,
+      "grad_norm": 0.27784302830696106,
+      "learning_rate": 5.536870991504044e-05,
+      "loss": 0.1135,
+      "step": 12300
+    },
+    {
+      "epoch": 321.83006535947715,
+      "grad_norm": 0.2698685824871063,
+      "learning_rate": 5.522080578625438e-05,
+      "loss": 0.1134,
+      "step": 12310
+    },
+    {
+      "epoch": 322.09150326797385,
+      "grad_norm": 0.3085765838623047,
+      "learning_rate": 5.507302408207542e-05,
+      "loss": 0.1138,
+      "step": 12320
+    },
+    {
+      "epoch": 322.3529411764706,
+      "grad_norm": 0.3435112237930298,
+      "learning_rate": 5.492536520653307e-05,
+      "loss": 0.1124,
+      "step": 12330
+    },
+    {
+      "epoch": 322.6143790849673,
+      "grad_norm": 0.3543654680252075,
+      "learning_rate": 5.4777829563321046e-05,
+      "loss": 0.1152,
+      "step": 12340
+    },
+    {
+      "epoch": 322.87581699346407,
+      "grad_norm": 0.2766146957874298,
+      "learning_rate": 5.463041755579619e-05,
+      "loss": 0.1144,
+      "step": 12350
+    },
+    {
+      "epoch": 323.1372549019608,
+      "grad_norm": 0.3165183961391449,
+      "learning_rate": 5.4483129586977386e-05,
+      "loss": 0.1154,
+      "step": 12360
+    },
+    {
+      "epoch": 323.39869281045753,
+      "grad_norm": 0.3891643285751343,
+      "learning_rate": 5.433596605954415e-05,
+      "loss": 0.1137,
+      "step": 12370
+    },
+    {
+      "epoch": 323.66013071895424,
+      "grad_norm": 0.28057435154914856,
+      "learning_rate": 5.4188927375836074e-05,
+      "loss": 0.1116,
+      "step": 12380
+    },
+    {
+      "epoch": 323.921568627451,
+      "grad_norm": 0.2857229709625244,
+      "learning_rate": 5.404201393785122e-05,
+      "loss": 0.1143,
+      "step": 12390
+    },
+    {
+      "epoch": 324.1830065359477,
+      "grad_norm": 0.31078898906707764,
+      "learning_rate": 5.389522614724536e-05,
+      "loss": 0.113,
+      "step": 12400
+    },
+    {
+      "epoch": 324.44444444444446,
+      "grad_norm": 0.2750803530216217,
+      "learning_rate": 5.374856440533078e-05,
+      "loss": 0.1129,
+      "step": 12410
+    },
+    {
+      "epoch": 324.70588235294116,
+      "grad_norm": 0.3001578748226166,
+      "learning_rate": 5.360202911307493e-05,
+      "loss": 0.1167,
+      "step": 12420
+    },
+    {
+      "epoch": 324.9673202614379,
+      "grad_norm": 0.2825237810611725,
+      "learning_rate": 5.345562067109984e-05,
+      "loss": 0.1131,
+      "step": 12430
+    },
+    {
+      "epoch": 325.2287581699346,
+      "grad_norm": 0.40779075026512146,
+      "learning_rate": 5.3309339479680485e-05,
+      "loss": 0.1165,
+      "step": 12440
+    },
+    {
+      "epoch": 325.4901960784314,
+      "grad_norm": 0.30739492177963257,
+      "learning_rate": 5.316318593874415e-05,
+      "loss": 0.1128,
+      "step": 12450
+    },
+    {
+      "epoch": 325.7516339869281,
+      "grad_norm": 0.2763489782810211,
+      "learning_rate": 5.301716044786902e-05,
+      "loss": 0.1109,
+      "step": 12460
+    },
+    {
+      "epoch": 326.01307189542484,
+      "grad_norm": 0.3952416479587555,
+      "learning_rate": 5.287126340628312e-05,
+      "loss": 0.1146,
+      "step": 12470
+    },
+    {
+      "epoch": 326.27450980392155,
+      "grad_norm": 0.36667054891586304,
+      "learning_rate": 5.2725495212863494e-05,
+      "loss": 0.1158,
+      "step": 12480
+    },
+    {
+      "epoch": 326.5359477124183,
+      "grad_norm": 0.2867089807987213,
+      "learning_rate": 5.25798562661348e-05,
+      "loss": 0.1112,
+      "step": 12490
+    },
+    {
+      "epoch": 326.797385620915,
+      "grad_norm": 0.39078038930892944,
+      "learning_rate": 5.2434346964268344e-05,
+      "loss": 0.1159,
+      "step": 12500
+    },
+    {
+      "epoch": 327.05882352941177,
+      "grad_norm": 0.30490151047706604,
+      "learning_rate": 5.2288967705081e-05,
+      "loss": 0.116,
+      "step": 12510
+    },
+    {
+      "epoch": 327.32026143790847,
+      "grad_norm": 0.2912214696407318,
+      "learning_rate": 5.214371888603409e-05,
+      "loss": 0.1097,
+      "step": 12520
+    },
+    {
+      "epoch": 327.58169934640523,
+      "grad_norm": 0.32746198773384094,
+      "learning_rate": 5.199860090423233e-05,
+      "loss": 0.1136,
+      "step": 12530
+    },
+    {
+      "epoch": 327.84313725490193,
+      "grad_norm": 0.48329511284828186,
+      "learning_rate": 5.185361415642283e-05,
+      "loss": 0.1163,
+      "step": 12540
+    },
+    {
+      "epoch": 328.1045751633987,
+      "grad_norm": 0.34038567543029785,
+      "learning_rate": 5.170875903899375e-05,
+      "loss": 0.1136,
+      "step": 12550
+    },
+    {
+      "epoch": 328.36601307189545,
+      "grad_norm": 0.3485356569290161,
+      "learning_rate": 5.1564035947973456e-05,
+      "loss": 0.1106,
+      "step": 12560
+    },
+    {
+      "epoch": 328.62745098039215,
+      "grad_norm": 0.3116021156311035,
+      "learning_rate": 5.141944527902932e-05,
+      "loss": 0.1147,
+      "step": 12570
+    },
+    {
+      "epoch": 328.8888888888889,
+      "grad_norm": 0.32443079352378845,
+      "learning_rate": 5.127498742746675e-05,
+      "loss": 0.1181,
+      "step": 12580
+    },
+    {
+      "epoch": 329.1503267973856,
+      "grad_norm": 0.28348246216773987,
+      "learning_rate": 5.113066278822807e-05,
+      "loss": 0.1125,
+      "step": 12590
+    },
+    {
+      "epoch": 329.4117647058824,
+      "grad_norm": 0.30532312393188477,
+      "learning_rate": 5.098647175589118e-05,
+      "loss": 0.1122,
+      "step": 12600
+    },
+    {
+      "epoch": 329.6732026143791,
+      "grad_norm": 0.37380146980285645,
+      "learning_rate": 5.084241472466897e-05,
+      "loss": 0.1149,
+      "step": 12610
+    },
+    {
+      "epoch": 329.93464052287584,
+      "grad_norm": 0.2732062339782715,
+      "learning_rate": 5.069849208840779e-05,
+      "loss": 0.1124,
+      "step": 12620
+    },
+    {
+      "epoch": 330.19607843137254,
+      "grad_norm": 0.3351595997810364,
+      "learning_rate": 5.055470424058666e-05,
+      "loss": 0.1137,
+      "step": 12630
+    },
+    {
+      "epoch": 330.4575163398693,
+      "grad_norm": 0.33784759044647217,
+      "learning_rate": 5.041105157431616e-05,
+      "loss": 0.114,
+      "step": 12640
+    },
+    {
+      "epoch": 330.718954248366,
+      "grad_norm": 0.30338507890701294,
+      "learning_rate": 5.026753448233703e-05,
+      "loss": 0.1135,
+      "step": 12650
+    },
+    {
+      "epoch": 330.98039215686276,
+      "grad_norm": 0.35373762249946594,
+      "learning_rate": 5.012415335701962e-05,
+      "loss": 0.1145,
+      "step": 12660
+    },
+    {
+      "epoch": 331.24183006535947,
+      "grad_norm": 0.3398035764694214,
+      "learning_rate": 4.99809085903624e-05,
+      "loss": 0.1119,
+      "step": 12670
+    },
+    {
+      "epoch": 331.5032679738562,
+      "grad_norm": 0.3283422291278839,
+      "learning_rate": 4.983780057399111e-05,
+      "loss": 0.1151,
+      "step": 12680
+    },
+    {
+      "epoch": 331.7647058823529,
+      "grad_norm": 0.30405548214912415,
+      "learning_rate": 4.9694829699157695e-05,
+      "loss": 0.1136,
+      "step": 12690
+    },
+    {
+      "epoch": 332.0261437908497,
+      "grad_norm": 0.3022817373275757,
+      "learning_rate": 4.9551996356738915e-05,
+      "loss": 0.1128,
+      "step": 12700
+    },
+    {
+      "epoch": 332.2875816993464,
+      "grad_norm": 0.3224383592605591,
+      "learning_rate": 4.940930093723578e-05,
+      "loss": 0.1129,
+      "step": 12710
+    },
+    {
+      "epoch": 332.54901960784315,
+      "grad_norm": 0.3512856662273407,
+      "learning_rate": 4.9266743830772034e-05,
+      "loss": 0.1125,
+      "step": 12720
+    },
+    {
+      "epoch": 332.81045751633985,
+      "grad_norm": 0.32787182927131653,
+      "learning_rate": 4.9124325427093455e-05,
+      "loss": 0.1132,
+      "step": 12730
+    },
+    {
+      "epoch": 333.0718954248366,
+      "grad_norm": 0.29392367601394653,
+      "learning_rate": 4.898204611556647e-05,
+      "loss": 0.1154,
+      "step": 12740
+    },
+    {
+      "epoch": 333.3333333333333,
+      "grad_norm": 0.30844447016716003,
+      "learning_rate": 4.883990628517725e-05,
+      "loss": 0.112,
+      "step": 12750
+    },
+    {
+      "epoch": 333.5947712418301,
+      "grad_norm": 0.38730964064598083,
+      "learning_rate": 4.869790632453075e-05,
+      "loss": 0.1178,
+      "step": 12760
+    },
+    {
+      "epoch": 333.8562091503268,
+      "grad_norm": 0.3621443808078766,
+      "learning_rate": 4.8556046621849346e-05,
+      "loss": 0.1114,
+      "step": 12770
+    },
+    {
+      "epoch": 334.11764705882354,
+      "grad_norm": 0.31759655475616455,
+      "learning_rate": 4.841432756497214e-05,
+      "loss": 0.1159,
+      "step": 12780
+    },
+    {
+      "epoch": 334.37908496732024,
+      "grad_norm": 0.2881384789943695,
+      "learning_rate": 4.827274954135358e-05,
+      "loss": 0.1123,
+      "step": 12790
+    },
+    {
+      "epoch": 334.640522875817,
+      "grad_norm": 0.3464260697364807,
+      "learning_rate": 4.813131293806253e-05,
+      "loss": 0.1125,
+      "step": 12800
+    },
+    {
+      "epoch": 334.9019607843137,
+      "grad_norm": 0.29048338532447815,
+      "learning_rate": 4.7990018141781344e-05,
+      "loss": 0.1154,
+      "step": 12810
+    },
+    {
+      "epoch": 335.16339869281046,
+      "grad_norm": 0.301886647939682,
+      "learning_rate": 4.7848865538804535e-05,
+      "loss": 0.1129,
+      "step": 12820
+    },
+    {
+      "epoch": 335.42483660130716,
+      "grad_norm": 0.3119642436504364,
+      "learning_rate": 4.770785551503798e-05,
+      "loss": 0.1118,
+      "step": 12830
+    },
+    {
+      "epoch": 335.6862745098039,
+      "grad_norm": 0.32431304454803467,
+      "learning_rate": 4.756698845599769e-05,
+      "loss": 0.1127,
+      "step": 12840
+    },
+    {
+      "epoch": 335.9477124183006,
+      "grad_norm": 0.41362830996513367,
+      "learning_rate": 4.7426264746808755e-05,
+      "loss": 0.1163,
+      "step": 12850
+    },
+    {
+      "epoch": 336.2091503267974,
+      "grad_norm": 0.2932046055793762,
+      "learning_rate": 4.728568477220453e-05,
+      "loss": 0.1129,
+      "step": 12860
+    },
+    {
+      "epoch": 336.47058823529414,
+      "grad_norm": 0.32170718908309937,
+      "learning_rate": 4.714524891652524e-05,
+      "loss": 0.1123,
+      "step": 12870
+    },
+    {
+      "epoch": 336.73202614379085,
+      "grad_norm": 0.32445117831230164,
+      "learning_rate": 4.7004957563717134e-05,
+      "loss": 0.1132,
+      "step": 12880
+    },
+    {
+      "epoch": 336.9934640522876,
+      "grad_norm": 0.39626502990722656,
+      "learning_rate": 4.686481109733146e-05,
+      "loss": 0.1167,
+      "step": 12890
+    },
+    {
+      "epoch": 337.2549019607843,
+      "grad_norm": 0.2989867925643921,
+      "learning_rate": 4.6724809900523256e-05,
+      "loss": 0.1099,
+      "step": 12900
+    },
+    {
+      "epoch": 337.51633986928107,
+      "grad_norm": 0.303310364484787,
+      "learning_rate": 4.658495435605051e-05,
+      "loss": 0.1151,
+      "step": 12910
+    },
+    {
+      "epoch": 337.77777777777777,
+      "grad_norm": 0.28980472683906555,
+      "learning_rate": 4.6445244846272916e-05,
+      "loss": 0.1143,
+      "step": 12920
+    },
+    {
+      "epoch": 338.03921568627453,
+      "grad_norm": 0.2834646999835968,
+      "learning_rate": 4.630568175315088e-05,
+      "loss": 0.1138,
+      "step": 12930
+    },
+    {
+      "epoch": 338.30065359477123,
+      "grad_norm": 0.2876114249229431,
+      "learning_rate": 4.6166265458244665e-05,
+      "loss": 0.1142,
+      "step": 12940
+    },
+    {
+      "epoch": 338.562091503268,
+      "grad_norm": 0.35517698526382446,
+      "learning_rate": 4.6026996342713e-05,
+      "loss": 0.1116,
+      "step": 12950
+    },
+    {
+      "epoch": 338.8235294117647,
+      "grad_norm": 0.310722678899765,
+      "learning_rate": 4.588787478731242e-05,
+      "loss": 0.1145,
+      "step": 12960
+    },
+    {
+      "epoch": 339.08496732026146,
+      "grad_norm": 0.2860526740550995,
+      "learning_rate": 4.574890117239592e-05,
+      "loss": 0.1144,
+      "step": 12970
+    },
+    {
+      "epoch": 339.34640522875816,
+      "grad_norm": 0.3224795162677765,
+      "learning_rate": 4.5610075877912e-05,
+      "loss": 0.1144,
+      "step": 12980
+    },
+    {
+      "epoch": 339.6078431372549,
+      "grad_norm": 0.3138015866279602,
+      "learning_rate": 4.5471399283403784e-05,
+      "loss": 0.1126,
+      "step": 12990
+    },
+    {
+      "epoch": 339.8692810457516,
+      "grad_norm": 0.3215549886226654,
+      "learning_rate": 4.533287176800772e-05,
+      "loss": 0.114,
+      "step": 13000
+    },
+    {
+      "epoch": 340.1307189542484,
+      "grad_norm": 0.2976275384426117,
+      "learning_rate": 4.5194493710452825e-05,
+      "loss": 0.1131,
+      "step": 13010
+    },
+    {
+      "epoch": 340.3921568627451,
+      "grad_norm": 0.3737821877002716,
+      "learning_rate": 4.505626548905938e-05,
+      "loss": 0.1141,
+      "step": 13020
+    },
+    {
+      "epoch": 340.65359477124184,
+      "grad_norm": 0.3438400328159332,
+      "learning_rate": 4.491818748173804e-05,
+      "loss": 0.1133,
+      "step": 13030
+    },
+    {
+      "epoch": 340.91503267973854,
+      "grad_norm": 0.34674420952796936,
+      "learning_rate": 4.478026006598885e-05,
+      "loss": 0.1134,
+      "step": 13040
+    },
+    {
+      "epoch": 341.1764705882353,
+      "grad_norm": 0.31675341725349426,
+      "learning_rate": 4.464248361890006e-05,
+      "loss": 0.112,
+      "step": 13050
+    },
+    {
+      "epoch": 341.437908496732,
+      "grad_norm": 0.2978508770465851,
+      "learning_rate": 4.4504858517147265e-05,
+      "loss": 0.1124,
+      "step": 13060
+    },
+    {
+      "epoch": 341.69934640522877,
+      "grad_norm": 0.32540374994277954,
+      "learning_rate": 4.43673851369922e-05,
+      "loss": 0.1104,
+      "step": 13070
+    },
+    {
+      "epoch": 341.96078431372547,
+      "grad_norm": 0.3203536570072174,
+      "learning_rate": 4.423006385428181e-05,
+      "loss": 0.119,
+      "step": 13080
+    },
+    {
+      "epoch": 342.22222222222223,
+      "grad_norm": 0.3339461386203766,
+      "learning_rate": 4.409289504444732e-05,
+      "loss": 0.1154,
+      "step": 13090
+    },
+    {
+      "epoch": 342.48366013071893,
+      "grad_norm": 0.30102136731147766,
+      "learning_rate": 4.3955879082502926e-05,
+      "loss": 0.1116,
+      "step": 13100
+    },
+    {
+      "epoch": 342.7450980392157,
+      "grad_norm": 0.31064385175704956,
+      "learning_rate": 4.381901634304512e-05,
+      "loss": 0.1145,
+      "step": 13110
+    },
+    {
+      "epoch": 343.0065359477124,
+      "grad_norm": 0.30831557512283325,
+      "learning_rate": 4.368230720025137e-05,
+      "loss": 0.1121,
+      "step": 13120
+    },
+    {
+      "epoch": 343.26797385620915,
+      "grad_norm": 0.31142985820770264,
+      "learning_rate": 4.35457520278792e-05,
+      "loss": 0.1136,
+      "step": 13130
+    },
+    {
+      "epoch": 343.52941176470586,
+      "grad_norm": 0.2939104735851288,
+      "learning_rate": 4.340935119926534e-05,
+      "loss": 0.1129,
+      "step": 13140
+    },
+    {
+      "epoch": 343.7908496732026,
+      "grad_norm": 0.30027836561203003,
+      "learning_rate": 4.327310508732437e-05,
+      "loss": 0.1144,
+      "step": 13150
+    },
+    {
+      "epoch": 344.0522875816994,
+      "grad_norm": 0.25800076127052307,
+      "learning_rate": 4.3137014064547965e-05,
+      "loss": 0.1144,
+      "step": 13160
+    },
+    {
+      "epoch": 344.3137254901961,
+      "grad_norm": 0.3547254204750061,
+      "learning_rate": 4.3001078503003825e-05,
+      "loss": 0.1115,
+      "step": 13170
+    },
+    {
+      "epoch": 344.57516339869284,
+      "grad_norm": 0.2877454459667206,
+      "learning_rate": 4.286529877433453e-05,
+      "loss": 0.1131,
+      "step": 13180
+    },
+    {
+      "epoch": 344.83660130718954,
+      "grad_norm": 0.36441680788993835,
+      "learning_rate": 4.272967524975673e-05,
+      "loss": 0.1127,
+      "step": 13190
+    },
+    {
+      "epoch": 345.0980392156863,
+      "grad_norm": 0.3238193988800049,
+      "learning_rate": 4.2594208300059946e-05,
+      "loss": 0.1185,
+      "step": 13200
+    },
+    {
+      "epoch": 345.359477124183,
+      "grad_norm": 0.2932418882846832,
+      "learning_rate": 4.245889829560559e-05,
+      "loss": 0.1134,
+      "step": 13210
+    },
+    {
+      "epoch": 345.62091503267976,
+      "grad_norm": 0.2501135766506195,
+      "learning_rate": 4.232374560632614e-05,
+      "loss": 0.108,
+      "step": 13220
+    },
+    {
+      "epoch": 345.88235294117646,
+      "grad_norm": 0.39940959215164185,
+      "learning_rate": 4.218875060172379e-05,
+      "loss": 0.1159,
+      "step": 13230
+    },
+    {
+      "epoch": 346.1437908496732,
+      "grad_norm": 0.30277755856513977,
+      "learning_rate": 4.2053913650869816e-05,
+      "loss": 0.1145,
+      "step": 13240
+    },
+    {
+      "epoch": 346.4052287581699,
+      "grad_norm": 0.2736481726169586,
+      "learning_rate": 4.191923512240327e-05,
+      "loss": 0.1114,
+      "step": 13250
+    },
+    {
+      "epoch": 346.6666666666667,
+      "grad_norm": 0.3326358199119568,
+      "learning_rate": 4.1784715384530035e-05,
+      "loss": 0.1131,
+      "step": 13260
+    },
+    {
+      "epoch": 346.9281045751634,
+      "grad_norm": 0.31444019079208374,
+      "learning_rate": 4.165035480502204e-05,
+      "loss": 0.1164,
+      "step": 13270
+    },
+    {
+      "epoch": 347.18954248366015,
+      "grad_norm": 0.3125884532928467,
+      "learning_rate": 4.1516153751215895e-05,
+      "loss": 0.1157,
+      "step": 13280
+    },
+    {
+      "epoch": 347.45098039215685,
+      "grad_norm": 0.3291109800338745,
+      "learning_rate": 4.138211259001222e-05,
+      "loss": 0.1151,
+      "step": 13290
+    },
+    {
+      "epoch": 347.7124183006536,
+      "grad_norm": 0.385786771774292,
+      "learning_rate": 4.1248231687874414e-05,
+      "loss": 0.1105,
+      "step": 13300
+    },
+    {
+      "epoch": 347.9738562091503,
+      "grad_norm": 0.32350867986679077,
+      "learning_rate": 4.1114511410827714e-05,
+      "loss": 0.1148,
+      "step": 13310
+    },
+    {
+      "epoch": 348.2352941176471,
+      "grad_norm": 0.2844417691230774,
+      "learning_rate": 4.098095212445831e-05,
+      "loss": 0.113,
+      "step": 13320
+    },
+    {
+      "epoch": 348.4967320261438,
+      "grad_norm": 0.298141747713089,
+      "learning_rate": 4.084755419391213e-05,
+      "loss": 0.1118,
+      "step": 13330
+    },
+    {
+      "epoch": 348.75816993464053,
+      "grad_norm": 0.2766071856021881,
+      "learning_rate": 4.071431798389408e-05,
+      "loss": 0.1134,
+      "step": 13340
+    },
+    {
+      "epoch": 349.01960784313724,
+      "grad_norm": 0.3159619867801666,
+      "learning_rate": 4.058124385866685e-05,
+      "loss": 0.1138,
+      "step": 13350
+    },
+    {
+      "epoch": 349.281045751634,
+      "grad_norm": 0.2848956286907196,
+      "learning_rate": 4.044833218204998e-05,
+      "loss": 0.1128,
+      "step": 13360
+    },
+    {
+      "epoch": 349.5424836601307,
+      "grad_norm": 0.28767091035842896,
+      "learning_rate": 4.031558331741897e-05,
+      "loss": 0.1114,
+      "step": 13370
+    },
+    {
+      "epoch": 349.80392156862746,
+      "grad_norm": 0.30374428629875183,
+      "learning_rate": 4.01829976277041e-05,
+      "loss": 0.1143,
+      "step": 13380
+    },
+    {
+      "epoch": 350.06535947712416,
+      "grad_norm": 0.2949487864971161,
+      "learning_rate": 4.005057547538964e-05,
+      "loss": 0.1147,
+      "step": 13390
+    },
+    {
+      "epoch": 350.3267973856209,
+      "grad_norm": 0.2627740502357483,
+      "learning_rate": 3.991831722251268e-05,
+      "loss": 0.1116,
+      "step": 13400
+    },
+    {
+      "epoch": 350.5882352941176,
+      "grad_norm": 0.3931275010108948,
+      "learning_rate": 3.978622323066217e-05,
+      "loss": 0.1147,
+      "step": 13410
+    },
+    {
+      "epoch": 350.8496732026144,
+      "grad_norm": 0.27783629298210144,
+      "learning_rate": 3.965429386097813e-05,
+      "loss": 0.1137,
+      "step": 13420
+    },
+    {
+      "epoch": 351.1111111111111,
+      "grad_norm": 0.29457733035087585,
+      "learning_rate": 3.952252947415038e-05,
+      "loss": 0.1134,
+      "step": 13430
+    },
+    {
+      "epoch": 351.37254901960785,
+      "grad_norm": 0.27213922142982483,
+      "learning_rate": 3.9390930430417696e-05,
+      "loss": 0.1098,
+      "step": 13440
+    },
+    {
+      "epoch": 351.63398692810455,
+      "grad_norm": 0.2786455452442169,
+      "learning_rate": 3.925949708956689e-05,
+      "loss": 0.1134,
+      "step": 13450
+    },
+    {
+      "epoch": 351.8954248366013,
+      "grad_norm": 0.326171338558197,
+      "learning_rate": 3.9128229810931626e-05,
+      "loss": 0.1145,
+      "step": 13460
+    },
+    {
+      "epoch": 352.15686274509807,
+      "grad_norm": 0.2971991002559662,
+      "learning_rate": 3.8997128953391727e-05,
+      "loss": 0.1141,
+      "step": 13470
+    },
+    {
+      "epoch": 352.41830065359477,
+      "grad_norm": 0.3068319261074066,
+      "learning_rate": 3.886619487537187e-05,
+      "loss": 0.1101,
+      "step": 13480
+    },
+    {
+      "epoch": 352.67973856209153,
+      "grad_norm": 0.285197377204895,
+      "learning_rate": 3.873542793484081e-05,
+      "loss": 0.1134,
+      "step": 13490
+    },
+    {
+      "epoch": 352.94117647058823,
+      "grad_norm": 0.30382540822029114,
+      "learning_rate": 3.860482848931042e-05,
+      "loss": 0.1161,
+      "step": 13500
+    },
+    {
+      "epoch": 353.202614379085,
+      "grad_norm": 0.32687294483184814,
+      "learning_rate": 3.847439689583454e-05,
+      "loss": 0.1113,
+      "step": 13510
+    },
+    {
+      "epoch": 353.4640522875817,
+      "grad_norm": 0.27805617451667786,
+      "learning_rate": 3.834413351100823e-05,
+      "loss": 0.1139,
+      "step": 13520
+    },
+    {
+      "epoch": 353.72549019607845,
+      "grad_norm": 0.33675503730773926,
+      "learning_rate": 3.821403869096658e-05,
+      "loss": 0.1166,
+      "step": 13530
+    },
+    {
+      "epoch": 353.98692810457516,
+      "grad_norm": 0.3418455123901367,
+      "learning_rate": 3.808411279138383e-05,
+      "loss": 0.1113,
+      "step": 13540
+    },
+    {
+      "epoch": 354.2483660130719,
+      "grad_norm": 0.2951265275478363,
+      "learning_rate": 3.7954356167472485e-05,
+      "loss": 0.1149,
+      "step": 13550
+    },
+    {
+      "epoch": 354.5098039215686,
+      "grad_norm": 0.3082883358001709,
+      "learning_rate": 3.782476917398213e-05,
+      "loss": 0.1125,
+      "step": 13560
+    },
+    {
+      "epoch": 354.7712418300654,
+      "grad_norm": 0.30738532543182373,
+      "learning_rate": 3.7695352165198774e-05,
+      "loss": 0.1139,
+      "step": 13570
+    },
+    {
+      "epoch": 355.0326797385621,
+      "grad_norm": 0.28504759073257446,
+      "learning_rate": 3.7566105494943435e-05,
+      "loss": 0.1122,
+      "step": 13580
+    },
+    {
+      "epoch": 355.29411764705884,
+      "grad_norm": 0.2820607125759125,
+      "learning_rate": 3.743702951657163e-05,
+      "loss": 0.1104,
+      "step": 13590
+    },
+    {
+      "epoch": 355.55555555555554,
+      "grad_norm": 0.38755741715431213,
+      "learning_rate": 3.730812458297222e-05,
+      "loss": 0.1149,
+      "step": 13600
+    },
+    {
+      "epoch": 355.8169934640523,
+      "grad_norm": 0.31920456886291504,
+      "learning_rate": 3.717939104656626e-05,
+      "loss": 0.1128,
+      "step": 13610
+    },
+    {
+      "epoch": 356.078431372549,
+      "grad_norm": 0.4119099974632263,
+      "learning_rate": 3.7050829259306466e-05,
+      "loss": 0.1142,
+      "step": 13620
+    },
+    {
+      "epoch": 356.33986928104576,
+      "grad_norm": 0.31791460514068604,
+      "learning_rate": 3.692243957267568e-05,
+      "loss": 0.1095,
+      "step": 13630
+    },
+    {
+      "epoch": 356.60130718954247,
+      "grad_norm": 0.24925312399864197,
+      "learning_rate": 3.679422233768651e-05,
+      "loss": 0.1179,
+      "step": 13640
+    },
+    {
+      "epoch": 356.8627450980392,
+      "grad_norm": 0.3441096246242523,
+      "learning_rate": 3.6666177904879994e-05,
+      "loss": 0.1141,
+      "step": 13650
+    },
+    {
+      "epoch": 357.12418300653593,
+      "grad_norm": 0.2764681577682495,
+      "learning_rate": 3.655108595056173e-05,
+      "loss": 0.1124,
+      "step": 13660
+    },
+    {
+      "epoch": 357.3856209150327,
+      "grad_norm": 0.29715630412101746,
+      "learning_rate": 3.6423370805949876e-05,
+      "loss": 0.1119,
+      "step": 13670
+    },
+    {
+      "epoch": 357.6470588235294,
+      "grad_norm": 0.282524049282074,
+      "learning_rate": 3.629582947741461e-05,
+      "loss": 0.1155,
+      "step": 13680
+    },
+    {
+      "epoch": 357.90849673202615,
+      "grad_norm": 0.3714596629142761,
+      "learning_rate": 3.616846231364902e-05,
+      "loss": 0.1121,
+      "step": 13690
+    },
+    {
+      "epoch": 358.16993464052285,
+      "grad_norm": 0.2508104145526886,
+      "learning_rate": 3.604126966287004e-05,
+      "loss": 0.1124,
+      "step": 13700
+    },
+    {
+      "epoch": 358.4313725490196,
+      "grad_norm": 0.30074048042297363,
+      "learning_rate": 3.591425187281756e-05,
+      "loss": 0.1143,
+      "step": 13710
+    },
+    {
+      "epoch": 358.6928104575163,
+      "grad_norm": 0.3082398474216461,
+      "learning_rate": 3.578740929075333e-05,
+      "loss": 0.1139,
+      "step": 13720
+    },
+    {
+      "epoch": 358.9542483660131,
+      "grad_norm": 0.3137090504169464,
+      "learning_rate": 3.5660742263460203e-05,
+      "loss": 0.1116,
+      "step": 13730
+    },
+    {
+      "epoch": 359.2156862745098,
+      "grad_norm": 0.33088791370391846,
+      "learning_rate": 3.553425113724088e-05,
+      "loss": 0.1108,
+      "step": 13740
+    },
+    {
+      "epoch": 359.47712418300654,
+      "grad_norm": 0.29915428161621094,
+      "learning_rate": 3.5407936257917326e-05,
+      "loss": 0.1142,
+      "step": 13750
+    },
+    {
+      "epoch": 359.73856209150324,
+      "grad_norm": 0.2895696461200714,
+      "learning_rate": 3.5281797970829635e-05,
+      "loss": 0.1125,
+      "step": 13760
+    },
+    {
+      "epoch": 360.0,
+      "grad_norm": 0.33440282940864563,
+      "learning_rate": 3.5155836620835006e-05,
+      "loss": 0.115,
+      "step": 13770
+    },
+    {
+      "epoch": 360.26143790849676,
+      "grad_norm": 0.3408808708190918,
+      "learning_rate": 3.5030052552307044e-05,
+      "loss": 0.1107,
+      "step": 13780
+    },
+    {
+      "epoch": 360.52287581699346,
+      "grad_norm": 0.35789361596107483,
+      "learning_rate": 3.490444610913447e-05,
+      "loss": 0.1147,
+      "step": 13790
+    },
+    {
+      "epoch": 360.7843137254902,
+      "grad_norm": 0.30055010318756104,
+      "learning_rate": 3.477901763472057e-05,
+      "loss": 0.1123,
+      "step": 13800
+    },
+    {
+      "epoch": 361.0457516339869,
+      "grad_norm": 0.3059723973274231,
+      "learning_rate": 3.465376747198203e-05,
+      "loss": 0.1129,
+      "step": 13810
+    },
+    {
+      "epoch": 361.3071895424837,
+      "grad_norm": 0.32120776176452637,
+      "learning_rate": 3.452869596334798e-05,
+      "loss": 0.1127,
+      "step": 13820
+    },
+    {
+      "epoch": 361.5686274509804,
+      "grad_norm": 0.3302242159843445,
+      "learning_rate": 3.440380345075915e-05,
+      "loss": 0.1121,
+      "step": 13830
+    },
+    {
+      "epoch": 361.83006535947715,
+      "grad_norm": 0.3632891774177551,
+      "learning_rate": 3.427909027566688e-05,
+      "loss": 0.1118,
+      "step": 13840
+    },
+    {
+      "epoch": 362.09150326797385,
+      "grad_norm": 0.26599353551864624,
+      "learning_rate": 3.415455677903224e-05,
+      "loss": 0.1149,
+      "step": 13850
+    },
+    {
+      "epoch": 362.3529411764706,
+      "grad_norm": 0.29619720578193665,
+      "learning_rate": 3.403020330132509e-05,
+      "loss": 0.1107,
+      "step": 13860
+    },
+    {
+      "epoch": 362.6143790849673,
+      "grad_norm": 0.3457578718662262,
+      "learning_rate": 3.3906030182523077e-05,
+      "loss": 0.1157,
+      "step": 13870
+    },
+    {
+      "epoch": 362.87581699346407,
+      "grad_norm": 0.33504214882850647,
+      "learning_rate": 3.378203776211075e-05,
+      "loss": 0.1117,
+      "step": 13880
+    },
+    {
+      "epoch": 363.1372549019608,
+      "grad_norm": 0.3943881392478943,
+      "learning_rate": 3.365822637907862e-05,
+      "loss": 0.1158,
+      "step": 13890
+    },
+    {
+      "epoch": 363.39869281045753,
+      "grad_norm": 0.4457724392414093,
+      "learning_rate": 3.353459637192231e-05,
+      "loss": 0.1122,
+      "step": 13900
+    },
+    {
+      "epoch": 363.66013071895424,
+      "grad_norm": 0.3329622745513916,
+      "learning_rate": 3.341114807864158e-05,
+      "loss": 0.1119,
+      "step": 13910
+    },
+    {
+      "epoch": 363.921568627451,
+      "grad_norm": 0.3475494682788849,
+      "learning_rate": 3.328788183673932e-05,
+      "loss": 0.1142,
+      "step": 13920
+    },
+    {
+      "epoch": 364.1830065359477,
+      "grad_norm": 0.2990402281284332,
+      "learning_rate": 3.316479798322072e-05,
+      "loss": 0.1116,
+      "step": 13930
+    },
+    {
+      "epoch": 364.44444444444446,
+      "grad_norm": 0.24483919143676758,
+      "learning_rate": 3.3041896854592305e-05,
+      "loss": 0.1128,
+      "step": 13940
+    },
+    {
+      "epoch": 364.70588235294116,
+      "grad_norm": 0.31257379055023193,
+      "learning_rate": 3.2919178786861104e-05,
+      "loss": 0.1143,
+      "step": 13950
+    },
+    {
+      "epoch": 364.9673202614379,
+      "grad_norm": 0.3180519938468933,
+      "learning_rate": 3.279664411553368e-05,
+      "loss": 0.1138,
+      "step": 13960
+    },
+    {
+      "epoch": 365.2287581699346,
+      "grad_norm": 0.32743024826049805,
+      "learning_rate": 3.267429317561504e-05,
+      "loss": 0.1112,
+      "step": 13970
+    },
+    {
+      "epoch": 365.4901960784314,
+      "grad_norm": 0.3621465861797333,
+      "learning_rate": 3.2552126301608043e-05,
+      "loss": 0.1136,
+      "step": 13980
+    },
+    {
+      "epoch": 365.7516339869281,
+      "grad_norm": 0.3021661639213562,
+      "learning_rate": 3.243014382751224e-05,
+      "loss": 0.1142,
+      "step": 13990
+    },
+    {
+      "epoch": 366.01307189542484,
+      "grad_norm": 0.4712045192718506,
+      "learning_rate": 3.230834608682305e-05,
+      "loss": 0.1143,
+      "step": 14000
+    },
+    {
+      "epoch": 366.27450980392155,
+      "grad_norm": 0.28367525339126587,
+      "learning_rate": 3.218673341253092e-05,
+      "loss": 0.1144,
+      "step": 14010
+    },
+    {
+      "epoch": 366.5359477124183,
+      "grad_norm": 0.36586108803749084,
+      "learning_rate": 3.206530613712014e-05,
+      "loss": 0.1121,
+      "step": 14020
+    },
+    {
+      "epoch": 366.797385620915,
+      "grad_norm": 0.2866863012313843,
+      "learning_rate": 3.194406459256833e-05,
+      "loss": 0.1137,
+      "step": 14030
+    },
+    {
+      "epoch": 367.05882352941177,
+      "grad_norm": 0.31462156772613525,
+      "learning_rate": 3.182300911034518e-05,
+      "loss": 0.1129,
+      "step": 14040
+    },
+    {
+      "epoch": 367.32026143790847,
+      "grad_norm": 0.2943307161331177,
+      "learning_rate": 3.17021400214118e-05,
+      "loss": 0.1101,
+      "step": 14050
+    },
+    {
+      "epoch": 367.58169934640523,
+      "grad_norm": 0.31919369101524353,
+      "learning_rate": 3.158145765621971e-05,
+      "loss": 0.1132,
+      "step": 14060
+    },
+    {
+      "epoch": 367.84313725490193,
+      "grad_norm": 0.31484705209732056,
+      "learning_rate": 3.1460962344709774e-05,
+      "loss": 0.1127,
+      "step": 14070
+    },
+    {
+      "epoch": 368.1045751633987,
+      "grad_norm": 0.3020099997520447,
+      "learning_rate": 3.1340654416311656e-05,
+      "loss": 0.1153,
+      "step": 14080
+    },
+    {
+      "epoch": 368.36601307189545,
+      "grad_norm": 0.29994770884513855,
+      "learning_rate": 3.1220534199942585e-05,
+      "loss": 0.113,
+      "step": 14090
+    },
+    {
+      "epoch": 368.62745098039215,
+      "grad_norm": 0.28115540742874146,
+      "learning_rate": 3.1100602024006707e-05,
+      "loss": 0.1115,
+      "step": 14100
+    },
+    {
+      "epoch": 368.8888888888889,
+      "grad_norm": 0.37485307455062866,
+      "learning_rate": 3.098085821639398e-05,
+      "loss": 0.1172,
+      "step": 14110
+    },
+    {
+      "epoch": 369.1503267973856,
+      "grad_norm": 0.3369976282119751,
+      "learning_rate": 3.086130310447937e-05,
+      "loss": 0.11,
+      "step": 14120
+    },
+    {
+      "epoch": 369.4117647058824,
+      "grad_norm": 0.2802003026008606,
+      "learning_rate": 3.074193701512204e-05,
+      "loss": 0.1113,
+      "step": 14130
+    },
+    {
+      "epoch": 369.6732026143791,
+      "grad_norm": 0.3496047854423523,
+      "learning_rate": 3.0622760274664275e-05,
+      "loss": 0.1167,
+      "step": 14140
+    },
+    {
+      "epoch": 369.93464052287584,
+      "grad_norm": 0.2986941933631897,
+      "learning_rate": 3.0503773208930787e-05,
+      "loss": 0.1118,
+      "step": 14150
+    },
+    {
+      "epoch": 370.19607843137254,
+      "grad_norm": 0.2768370807170868,
+      "learning_rate": 3.038497614322763e-05,
+      "loss": 0.1123,
+      "step": 14160
+    },
+    {
+      "epoch": 370.4575163398693,
+      "grad_norm": 0.3114430010318756,
+      "learning_rate": 3.0266369402341433e-05,
+      "loss": 0.1152,
+      "step": 14170
+    },
+    {
+      "epoch": 370.718954248366,
+      "grad_norm": 0.30271750688552856,
+      "learning_rate": 3.0147953310538546e-05,
+      "loss": 0.1107,
+      "step": 14180
+    },
+    {
+      "epoch": 370.98039215686276,
+      "grad_norm": 0.33484479784965515,
+      "learning_rate": 3.0029728191563977e-05,
+      "loss": 0.112,
+      "step": 14190
+    },
+    {
+      "epoch": 371.24183006535947,
+      "grad_norm": 0.428725004196167,
+      "learning_rate": 2.9911694368640764e-05,
+      "loss": 0.1123,
+      "step": 14200
+    },
+    {
+      "epoch": 371.5032679738562,
+      "grad_norm": 0.3206990361213684,
+      "learning_rate": 2.9793852164468826e-05,
+      "loss": 0.1128,
+      "step": 14210
+    },
+    {
+      "epoch": 371.7647058823529,
+      "grad_norm": 0.3497907817363739,
+      "learning_rate": 2.9676201901224233e-05,
+      "loss": 0.1133,
+      "step": 14220
+    },
+    {
+      "epoch": 372.0261437908497,
+      "grad_norm": 0.3005002439022064,
+      "learning_rate": 2.955874390055836e-05,
+      "loss": 0.1127,
+      "step": 14230
+    },
+    {
+      "epoch": 372.2875816993464,
+      "grad_norm": 0.2785640358924866,
+      "learning_rate": 2.9441478483596862e-05,
+      "loss": 0.1137,
+      "step": 14240
+    },
+    {
+      "epoch": 372.54901960784315,
+      "grad_norm": 0.33614808320999146,
+      "learning_rate": 2.9324405970938906e-05,
+      "loss": 0.1105,
+      "step": 14250
+    },
+    {
+      "epoch": 372.81045751633985,
+      "grad_norm": 0.2854107618331909,
+      "learning_rate": 2.9207526682656306e-05,
+      "loss": 0.1153,
+      "step": 14260
+    },
+    {
+      "epoch": 373.0718954248366,
+      "grad_norm": 0.25472962856292725,
+      "learning_rate": 2.909084093829252e-05,
+      "loss": 0.1101,
+      "step": 14270
+    },
+    {
+      "epoch": 373.3333333333333,
+      "grad_norm": 0.34902864694595337,
+      "learning_rate": 2.897434905686198e-05,
+      "loss": 0.1105,
+      "step": 14280
+    },
+    {
+      "epoch": 373.5947712418301,
+      "grad_norm": 0.2810913622379303,
+      "learning_rate": 2.8858051356849014e-05,
+      "loss": 0.1148,
+      "step": 14290
+    },
+    {
+      "epoch": 373.8562091503268,
+      "grad_norm": 0.34702831506729126,
+      "learning_rate": 2.8741948156207056e-05,
+      "loss": 0.1137,
+      "step": 14300
+    },
+    {
+      "epoch": 374.11764705882354,
+      "grad_norm": 0.31400611996650696,
+      "learning_rate": 2.8626039772357882e-05,
+      "loss": 0.1122,
+      "step": 14310
+    },
+    {
+      "epoch": 374.37908496732024,
+      "grad_norm": 0.3254755139350891,
+      "learning_rate": 2.8510326522190545e-05,
+      "loss": 0.1113,
+      "step": 14320
+    },
+    {
+      "epoch": 374.640522875817,
+      "grad_norm": 0.31944742798805237,
+      "learning_rate": 2.8394808722060696e-05,
+      "loss": 0.1145,
+      "step": 14330
+    },
+    {
+      "epoch": 374.9019607843137,
+      "grad_norm": 0.30928805470466614,
+      "learning_rate": 2.8279486687789558e-05,
+      "loss": 0.1135,
+      "step": 14340
+    },
+    {
+      "epoch": 375.16339869281046,
+      "grad_norm": 0.3214901089668274,
+      "learning_rate": 2.8164360734663142e-05,
+      "loss": 0.111,
+      "step": 14350
+    },
+    {
+      "epoch": 375.42483660130716,
+      "grad_norm": 0.2930542230606079,
+      "learning_rate": 2.8049431177431486e-05,
+      "loss": 0.1139,
+      "step": 14360
+    },
+    {
+      "epoch": 375.6862745098039,
+      "grad_norm": 0.3097141683101654,
+      "learning_rate": 2.7934698330307518e-05,
+      "loss": 0.1127,
+      "step": 14370
+    },
+    {
+      "epoch": 375.9477124183006,
+      "grad_norm": 0.33316555619239807,
+      "learning_rate": 2.782016250696655e-05,
+      "loss": 0.1139,
+      "step": 14380
+    },
+    {
+      "epoch": 376.2091503267974,
+      "grad_norm": 0.3537690043449402,
+      "learning_rate": 2.77058240205451e-05,
+      "loss": 0.1126,
+      "step": 14390
+    },
+    {
+      "epoch": 376.47058823529414,
+      "grad_norm": 0.35879310965538025,
+      "learning_rate": 2.7591683183640215e-05,
+      "loss": 0.1145,
+      "step": 14400
+    },
+    {
+      "epoch": 376.73202614379085,
+      "grad_norm": 0.3203032314777374,
+      "learning_rate": 2.7477740308308618e-05,
+      "loss": 0.1093,
+      "step": 14410
+    },
+    {
+      "epoch": 376.9934640522876,
+      "grad_norm": 0.3569747507572174,
+      "learning_rate": 2.7363995706065737e-05,
+      "loss": 0.1143,
+      "step": 14420
+    },
+    {
+      "epoch": 377.2549019607843,
+      "grad_norm": 0.33088165521621704,
+      "learning_rate": 2.7250449687885028e-05,
+      "loss": 0.1112,
+      "step": 14430
+    },
+    {
+      "epoch": 377.51633986928107,
+      "grad_norm": 0.4784368872642517,
+      "learning_rate": 2.7137102564196937e-05,
+      "loss": 0.1136,
+      "step": 14440
+    },
+    {
+      "epoch": 377.77777777777777,
+      "grad_norm": 0.39264166355133057,
+      "learning_rate": 2.702395464488814e-05,
+      "loss": 0.115,
+      "step": 14450
+    },
+    {
+      "epoch": 378.03921568627453,
+      "grad_norm": 0.3187940716743469,
+      "learning_rate": 2.6911006239300794e-05,
+      "loss": 0.1105,
+      "step": 14460
+    },
+    {
+      "epoch": 378.30065359477123,
+      "grad_norm": 0.30897989869117737,
+      "learning_rate": 2.6798257656231464e-05,
+      "loss": 0.1127,
+      "step": 14470
+    },
+    {
+      "epoch": 378.562091503268,
+      "grad_norm": 0.40761929750442505,
+      "learning_rate": 2.668570920393052e-05,
+      "loss": 0.1139,
+      "step": 14480
+    },
+    {
+      "epoch": 378.8235294117647,
+      "grad_norm": 0.32026803493499756,
+      "learning_rate": 2.657336119010112e-05,
+      "loss": 0.1136,
+      "step": 14490
+    },
+    {
+      "epoch": 379.08496732026146,
+      "grad_norm": 0.30582597851753235,
+      "learning_rate": 2.646121392189841e-05,
+      "loss": 0.112,
+      "step": 14500
+    },
+    {
+      "epoch": 379.34640522875816,
+      "grad_norm": 0.2957599461078644,
+      "learning_rate": 2.6349267705928793e-05,
+      "loss": 0.1105,
+      "step": 14510
+    },
+    {
+      "epoch": 379.6078431372549,
+      "grad_norm": 0.2888094484806061,
+      "learning_rate": 2.623752284824893e-05,
+      "loss": 0.1119,
+      "step": 14520
+    },
+    {
+      "epoch": 379.8692810457516,
+      "grad_norm": 0.32123002409935,
+      "learning_rate": 2.6125979654364952e-05,
+      "loss": 0.1138,
+      "step": 14530
+    },
+    {
+      "epoch": 380.1307189542484,
+      "grad_norm": 0.2851218283176422,
+      "learning_rate": 2.601463842923175e-05,
+      "loss": 0.1113,
+      "step": 14540
+    },
+    {
+      "epoch": 380.3921568627451,
+      "grad_norm": 0.31959566473960876,
+      "learning_rate": 2.5903499477251936e-05,
+      "loss": 0.1136,
+      "step": 14550
+    },
+    {
+      "epoch": 380.65359477124184,
+      "grad_norm": 0.32848477363586426,
+      "learning_rate": 2.5792563102275213e-05,
+      "loss": 0.1107,
+      "step": 14560
+    },
+    {
+      "epoch": 380.91503267973854,
+      "grad_norm": 0.31065839529037476,
+      "learning_rate": 2.568182960759735e-05,
+      "loss": 0.1144,
+      "step": 14570
+    },
+    {
+      "epoch": 381.1764705882353,
+      "grad_norm": 0.3056698739528656,
+      "learning_rate": 2.5571299295959496e-05,
+      "loss": 0.1127,
+      "step": 14580
+    },
+    {
+      "epoch": 381.437908496732,
+      "grad_norm": 0.29230251908302307,
+      "learning_rate": 2.546097246954734e-05,
+      "loss": 0.1102,
+      "step": 14590
+    },
+    {
+      "epoch": 381.69934640522877,
+      "grad_norm": 0.3203261196613312,
+      "learning_rate": 2.5350849429990152e-05,
+      "loss": 0.113,
+      "step": 14600
+    },
+    {
+      "epoch": 381.96078431372547,
+      "grad_norm": 0.406775563955307,
+      "learning_rate": 2.524093047836018e-05,
+      "loss": 0.1152,
+      "step": 14610
+    },
+    {
+      "epoch": 382.22222222222223,
+      "grad_norm": 0.2875988781452179,
+      "learning_rate": 2.5131215915171624e-05,
+      "loss": 0.113,
+      "step": 14620
+    },
+    {
+      "epoch": 382.48366013071893,
+      "grad_norm": 0.29914501309394836,
+      "learning_rate": 2.5021706040379854e-05,
+      "loss": 0.1104,
+      "step": 14630
+    },
+    {
+      "epoch": 382.7450980392157,
+      "grad_norm": 0.3158552944660187,
+      "learning_rate": 2.4912401153380772e-05,
+      "loss": 0.1152,
+      "step": 14640
+    },
+    {
+      "epoch": 383.0065359477124,
+      "grad_norm": 0.45473095774650574,
+      "learning_rate": 2.4803301553009694e-05,
+      "loss": 0.112,
+      "step": 14650
+    },
+    {
+      "epoch": 383.26797385620915,
+      "grad_norm": 0.32687652111053467,
+      "learning_rate": 2.4694407537540808e-05,
+      "loss": 0.1119,
+      "step": 14660
+    },
+    {
+      "epoch": 383.52941176470586,
+      "grad_norm": 0.3493711054325104,
+      "learning_rate": 2.4585719404686192e-05,
+      "loss": 0.1152,
+      "step": 14670
+    },
+    {
+      "epoch": 383.7908496732026,
+      "grad_norm": 0.3303303122520447,
+      "learning_rate": 2.4477237451595004e-05,
+      "loss": 0.1105,
+      "step": 14680
+    },
+    {
+      "epoch": 384.0522875816994,
+      "grad_norm": 0.29483675956726074,
+      "learning_rate": 2.436896197485282e-05,
+      "loss": 0.1136,
+      "step": 14690
+    },
+    {
+      "epoch": 384.3137254901961,
+      "grad_norm": 0.28735411167144775,
+      "learning_rate": 2.42608932704806e-05,
+      "loss": 0.1104,
+      "step": 14700
+    },
+    {
+      "epoch": 384.57516339869284,
+      "grad_norm": 0.30859243869781494,
+      "learning_rate": 2.415303163393412e-05,
+      "loss": 0.1133,
+      "step": 14710
+    },
+    {
+      "epoch": 384.83660130718954,
+      "grad_norm": 0.32482388615608215,
+      "learning_rate": 2.404537736010295e-05,
+      "loss": 0.1131,
+      "step": 14720
+    },
+    {
+      "epoch": 385.0980392156863,
+      "grad_norm": 0.32856839895248413,
+      "learning_rate": 2.3937930743309723e-05,
+      "loss": 0.113,
+      "step": 14730
+    },
+    {
+      "epoch": 385.359477124183,
+      "grad_norm": 0.3382302224636078,
+      "learning_rate": 2.3830692077309446e-05,
+      "loss": 0.111,
+      "step": 14740
+    },
+    {
+      "epoch": 385.62091503267976,
+      "grad_norm": 0.4078975319862366,
+      "learning_rate": 2.3723661655288487e-05,
+      "loss": 0.1147,
+      "step": 14750
+    },
+    {
+      "epoch": 385.88235294117646,
+      "grad_norm": 0.32090675830841064,
+      "learning_rate": 2.3616839769863984e-05,
+      "loss": 0.1121,
+      "step": 14760
+    },
+    {
+      "epoch": 386.1437908496732,
+      "grad_norm": 0.3056178092956543,
+      "learning_rate": 2.351022671308287e-05,
+      "loss": 0.1116,
+      "step": 14770
+    },
+    {
+      "epoch": 386.4052287581699,
+      "grad_norm": 0.33630117774009705,
+      "learning_rate": 2.3403822776421135e-05,
+      "loss": 0.113,
+      "step": 14780
+    },
+    {
+      "epoch": 386.6666666666667,
+      "grad_norm": 0.3036194145679474,
+      "learning_rate": 2.3297628250783154e-05,
+      "loss": 0.1124,
+      "step": 14790
+    },
+    {
+      "epoch": 386.9281045751634,
+      "grad_norm": 0.35577064752578735,
+      "learning_rate": 2.3191643426500653e-05,
+      "loss": 0.113,
+      "step": 14800
+    },
+    {
+      "epoch": 387.18954248366015,
+      "grad_norm": 0.3300045430660248,
+      "learning_rate": 2.3085868593332073e-05,
+      "loss": 0.1126,
+      "step": 14810
+    },
+    {
+      "epoch": 387.45098039215685,
+      "grad_norm": 0.28813156485557556,
+      "learning_rate": 2.298030404046183e-05,
+      "loss": 0.1106,
+      "step": 14820
+    },
+    {
+      "epoch": 387.7124183006536,
+      "grad_norm": 0.40409666299819946,
+      "learning_rate": 2.2874950056499324e-05,
+      "loss": 0.118,
+      "step": 14830
+    },
+    {
+      "epoch": 387.9738562091503,
+      "grad_norm": 0.3458053171634674,
+      "learning_rate": 2.2769806929478377e-05,
+      "loss": 0.1093,
+      "step": 14840
+    },
+    {
+      "epoch": 388.2352941176471,
+      "grad_norm": 0.3175845742225647,
+      "learning_rate": 2.266487494685625e-05,
+      "loss": 0.1122,
+      "step": 14850
+    },
+    {
+      "epoch": 388.4967320261438,
+      "grad_norm": 0.31592434644699097,
+      "learning_rate": 2.2560154395512967e-05,
+      "loss": 0.115,
+      "step": 14860
+    },
+    {
+      "epoch": 388.75816993464053,
+      "grad_norm": 0.31762829422950745,
+      "learning_rate": 2.245564556175056e-05,
+      "loss": 0.1127,
+      "step": 14870
+    },
+    {
+      "epoch": 389.01960784313724,
+      "grad_norm": 0.27463746070861816,
+      "learning_rate": 2.235134873129213e-05,
+      "loss": 0.1095,
+      "step": 14880
+    },
+    {
+      "epoch": 389.281045751634,
+      "grad_norm": 0.2556680738925934,
+      "learning_rate": 2.2247264189281304e-05,
+      "loss": 0.1127,
+      "step": 14890
+    },
+    {
+      "epoch": 389.5424836601307,
+      "grad_norm": 0.3717378079891205,
+      "learning_rate": 2.214339222028119e-05,
+      "loss": 0.1127,
+      "step": 14900
+    },
+    {
+      "epoch": 389.80392156862746,
+      "grad_norm": 0.33163613080978394,
+      "learning_rate": 2.2039733108273774e-05,
+      "loss": 0.1129,
+      "step": 14910
+    },
+    {
+      "epoch": 390.06535947712416,
+      "grad_norm": 0.33852529525756836,
+      "learning_rate": 2.193628713665916e-05,
+      "loss": 0.1126,
+      "step": 14920
+    },
+    {
+      "epoch": 390.3267973856209,
+      "grad_norm": 0.38463908433914185,
+      "learning_rate": 2.183305458825464e-05,
+      "loss": 0.1107,
+      "step": 14930
+    },
+    {
+      "epoch": 390.5882352941176,
+      "grad_norm": 0.29137659072875977,
+      "learning_rate": 2.1730035745294098e-05,
+      "loss": 0.1135,
+      "step": 14940
+    },
+    {
+      "epoch": 390.8496732026144,
+      "grad_norm": 0.3932526111602783,
+      "learning_rate": 2.1627230889427096e-05,
+      "loss": 0.1138,
+      "step": 14950
+    },
+    {
+      "epoch": 391.1111111111111,
+      "grad_norm": 0.30873677134513855,
+      "learning_rate": 2.1524640301718167e-05,
+      "loss": 0.1129,
+      "step": 14960
+    },
+    {
+      "epoch": 391.37254901960785,
+      "grad_norm": 0.28704315423965454,
+      "learning_rate": 2.1422264262646097e-05,
+      "loss": 0.1134,
+      "step": 14970
+    },
+    {
+      "epoch": 391.63398692810455,
+      "grad_norm": 0.3116617798805237,
+      "learning_rate": 2.1320103052103024e-05,
+      "loss": 0.1113,
+      "step": 14980
+    },
+    {
+      "epoch": 391.8954248366013,
+      "grad_norm": 0.3618609607219696,
+      "learning_rate": 2.1218156949393853e-05,
+      "loss": 0.1132,
+      "step": 14990
+    },
+    {
+      "epoch": 392.15686274509807,
+      "grad_norm": 0.30935177206993103,
+      "learning_rate": 2.111642623323531e-05,
+      "loss": 0.112,
+      "step": 15000
+    },
+    {
+      "epoch": 392.41830065359477,
+      "grad_norm": 0.32789334654808044,
+      "learning_rate": 2.1014911181755247e-05,
+      "loss": 0.1127,
+      "step": 15010
+    },
+    {
+      "epoch": 392.67973856209153,
+      "grad_norm": 0.3063357472419739,
+      "learning_rate": 2.0913612072492006e-05,
+      "loss": 0.1144,
+      "step": 15020
+    },
+    {
+      "epoch": 392.94117647058823,
+      "grad_norm": 0.3251248896121979,
+      "learning_rate": 2.0812529182393424e-05,
+      "loss": 0.1107,
+      "step": 15030
+    },
+    {
+      "epoch": 393.202614379085,
+      "grad_norm": 0.31256797909736633,
+      "learning_rate": 2.071166278781632e-05,
+      "loss": 0.11,
+      "step": 15040
+    },
+    {
+      "epoch": 393.4640522875817,
+      "grad_norm": 0.29990172386169434,
+      "learning_rate": 2.061101316452554e-05,
+      "loss": 0.1126,
+      "step": 15050
+    },
+    {
+      "epoch": 393.72549019607845,
+      "grad_norm": 0.34786394238471985,
+      "learning_rate": 2.0510580587693273e-05,
+      "loss": 0.1133,
+      "step": 15060
+    },
+    {
+      "epoch": 393.98692810457516,
+      "grad_norm": 0.33727478981018066,
+      "learning_rate": 2.0410365331898416e-05,
+      "loss": 0.1145,
+      "step": 15070
+    },
+    {
+      "epoch": 394.2483660130719,
+      "grad_norm": 0.31517842411994934,
+      "learning_rate": 2.0310367671125618e-05,
+      "loss": 0.1113,
+      "step": 15080
+    },
+    {
+      "epoch": 394.5098039215686,
+      "grad_norm": 0.36189743876457214,
+      "learning_rate": 2.021058787876464e-05,
+      "loss": 0.1151,
+      "step": 15090
+    },
+    {
+      "epoch": 394.7712418300654,
+      "grad_norm": 0.3063863515853882,
+      "learning_rate": 2.0111026227609675e-05,
+      "loss": 0.1124,
+      "step": 15100
+    },
+    {
+      "epoch": 395.0326797385621,
+      "grad_norm": 0.3760160207748413,
+      "learning_rate": 2.0011682989858427e-05,
+      "loss": 0.1107,
+      "step": 15110
+    },
+    {
+      "epoch": 395.29411764705884,
+      "grad_norm": 0.31587451696395874,
+      "learning_rate": 1.991255843711156e-05,
+      "loss": 0.1103,
+      "step": 15120
+    },
+    {
+      "epoch": 395.55555555555554,
+      "grad_norm": 0.3024122416973114,
+      "learning_rate": 1.98136528403718e-05,
+      "loss": 0.1093,
+      "step": 15130
+    },
+    {
+      "epoch": 395.8169934640523,
+      "grad_norm": 0.32457295060157776,
+      "learning_rate": 1.971496647004324e-05,
+      "loss": 0.1158,
+      "step": 15140
+    },
+    {
+      "epoch": 396.078431372549,
+      "grad_norm": 0.3319365978240967,
+      "learning_rate": 1.9616499595930692e-05,
+      "loss": 0.1151,
+      "step": 15150
+    },
+    {
+      "epoch": 396.33986928104576,
+      "grad_norm": 0.33728495240211487,
+      "learning_rate": 1.9518252487238797e-05,
+      "loss": 0.1118,
+      "step": 15160
+    },
+    {
+      "epoch": 396.60130718954247,
+      "grad_norm": 0.34540119767189026,
+      "learning_rate": 1.9420225412571435e-05,
+      "loss": 0.1121,
+      "step": 15170
+    },
+    {
+      "epoch": 396.8627450980392,
+      "grad_norm": 0.30559542775154114,
+      "learning_rate": 1.9322418639930863e-05,
+      "loss": 0.1125,
+      "step": 15180
+    },
+    {
+      "epoch": 397.12418300653593,
+      "grad_norm": 0.3552384376525879,
+      "learning_rate": 1.9224832436717045e-05,
+      "loss": 0.1133,
+      "step": 15190
+    },
+    {
+      "epoch": 397.3856209150327,
+      "grad_norm": 0.38219109177589417,
+      "learning_rate": 1.912746706972697e-05,
+      "loss": 0.1122,
+      "step": 15200
+    },
+    {
+      "epoch": 397.6470588235294,
+      "grad_norm": 0.38125985860824585,
+      "learning_rate": 1.90303228051538e-05,
+      "loss": 0.112,
+      "step": 15210
+    },
+    {
+      "epoch": 397.90849673202615,
+      "grad_norm": 0.3336564600467682,
+      "learning_rate": 1.89333999085863e-05,
+      "loss": 0.1127,
+      "step": 15220
+    },
+    {
+      "epoch": 398.16993464052285,
+      "grad_norm": 0.29126253724098206,
+      "learning_rate": 1.8836698645007888e-05,
+      "loss": 0.1115,
+      "step": 15230
+    },
+    {
+      "epoch": 398.4313725490196,
+      "grad_norm": 0.4003722667694092,
+      "learning_rate": 1.8740219278796167e-05,
+      "loss": 0.1117,
+      "step": 15240
+    },
+    {
+      "epoch": 398.6928104575163,
+      "grad_norm": 0.348296582698822,
+      "learning_rate": 1.8643962073722064e-05,
+      "loss": 0.1133,
+      "step": 15250
+    },
+    {
+      "epoch": 398.9542483660131,
+      "grad_norm": 0.37177759408950806,
+      "learning_rate": 1.854792729294905e-05,
+      "loss": 0.1121,
+      "step": 15260
+    },
+    {
+      "epoch": 399.2156862745098,
+      "grad_norm": 0.3493458032608032,
+      "learning_rate": 1.8452115199032638e-05,
+      "loss": 0.1116,
+      "step": 15270
+    },
+    {
+      "epoch": 399.47712418300654,
+      "grad_norm": 0.30324786901474,
+      "learning_rate": 1.835652605391931e-05,
+      "loss": 0.1116,
+      "step": 15280
+    },
+    {
+      "epoch": 399.73856209150324,
+      "grad_norm": 0.2836083471775055,
+      "learning_rate": 1.826116011894621e-05,
+      "loss": 0.111,
+      "step": 15290
+    },
+    {
+      "epoch": 400.0,
+      "grad_norm": 0.3711930513381958,
+      "learning_rate": 1.8166017654840184e-05,
+      "loss": 0.115,
+      "step": 15300
+    },
+    {
+      "epoch": 400.26143790849676,
+      "grad_norm": 0.3117862343788147,
+      "learning_rate": 1.8071098921717033e-05,
+      "loss": 0.1129,
+      "step": 15310
+    },
+    {
+      "epoch": 400.52287581699346,
+      "grad_norm": 0.3351728916168213,
+      "learning_rate": 1.797640417908104e-05,
+      "loss": 0.1117,
+      "step": 15320
+    },
+    {
+      "epoch": 400.7843137254902,
+      "grad_norm": 0.36364924907684326,
+      "learning_rate": 1.7881933685823905e-05,
+      "loss": 0.1146,
+      "step": 15330
+    },
+    {
+      "epoch": 401.0457516339869,
+      "grad_norm": 0.30129241943359375,
+      "learning_rate": 1.7787687700224397e-05,
+      "loss": 0.1101,
+      "step": 15340
+    },
+    {
+      "epoch": 401.3071895424837,
+      "grad_norm": 0.29976898431777954,
+      "learning_rate": 1.769366647994748e-05,
+      "loss": 0.1131,
+      "step": 15350
+    },
+    {
+      "epoch": 401.5686274509804,
+      "grad_norm": 0.28711673617362976,
+      "learning_rate": 1.7599870282043552e-05,
+      "loss": 0.1131,
+      "step": 15360
+    },
+    {
+      "epoch": 401.83006535947715,
+      "grad_norm": 0.32417982816696167,
+      "learning_rate": 1.750629936294782e-05,
+      "loss": 0.112,
+      "step": 15370
+    },
+    {
+      "epoch": 402.09150326797385,
+      "grad_norm": 0.2811816930770874,
+      "learning_rate": 1.7412953978479595e-05,
+      "loss": 0.1132,
+      "step": 15380
+    },
+    {
+      "epoch": 402.3529411764706,
+      "grad_norm": 0.3149554431438446,
+      "learning_rate": 1.7319834383841616e-05,
+      "loss": 0.11,
+      "step": 15390
+    },
+    {
+      "epoch": 402.6143790849673,
+      "grad_norm": 0.3484453856945038,
+      "learning_rate": 1.7226940833619322e-05,
+      "loss": 0.111,
+      "step": 15400
+    },
+    {
+      "epoch": 402.87581699346407,
+      "grad_norm": 0.3713397681713104,
+      "learning_rate": 1.7134273581780113e-05,
+      "loss": 0.1151,
+      "step": 15410
+    },
+    {
+      "epoch": 403.1372549019608,
+      "grad_norm": 0.36278796195983887,
+      "learning_rate": 1.7041832881672703e-05,
+      "loss": 0.1136,
+      "step": 15420
+    },
+    {
+      "epoch": 403.39869281045753,
+      "grad_norm": 0.29296210408210754,
+      "learning_rate": 1.6949618986026416e-05,
+      "loss": 0.1099,
+      "step": 15430
+    },
+    {
+      "epoch": 403.66013071895424,
+      "grad_norm": 0.32640260457992554,
+      "learning_rate": 1.6857632146950564e-05,
+      "loss": 0.1118,
+      "step": 15440
+    },
+    {
+      "epoch": 403.921568627451,
+      "grad_norm": 0.3738747537136078,
+      "learning_rate": 1.6765872615933677e-05,
+      "loss": 0.1151,
+      "step": 15450
+    },
+    {
+      "epoch": 404.1830065359477,
+      "grad_norm": 0.3174304962158203,
+      "learning_rate": 1.6674340643842733e-05,
+      "loss": 0.1092,
+      "step": 15460
+    },
+    {
+      "epoch": 404.44444444444446,
+      "grad_norm": 0.3067731261253357,
+      "learning_rate": 1.6583036480922697e-05,
+      "loss": 0.1133,
+      "step": 15470
+    },
+    {
+      "epoch": 404.70588235294116,
+      "grad_norm": 0.3003155589103699,
+      "learning_rate": 1.6491960376795635e-05,
+      "loss": 0.1139,
+      "step": 15480
+    },
+    {
+      "epoch": 404.9673202614379,
+      "grad_norm": 0.3668583631515503,
+      "learning_rate": 1.6401112580460167e-05,
+      "loss": 0.1122,
+      "step": 15490
+    },
+    {
+      "epoch": 405.2287581699346,
+      "grad_norm": 0.3048345148563385,
+      "learning_rate": 1.6310493340290723e-05,
+      "loss": 0.1122,
+      "step": 15500
+    },
+    {
+      "epoch": 405.4901960784314,
+      "grad_norm": 0.308072566986084,
+      "learning_rate": 1.622010290403677e-05,
+      "loss": 0.1172,
+      "step": 15510
+    },
+    {
+      "epoch": 405.7516339869281,
+      "grad_norm": 0.3233634829521179,
+      "learning_rate": 1.6129941518822366e-05,
+      "loss": 0.1103,
+      "step": 15520
+    },
+    {
+      "epoch": 406.01307189542484,
+      "grad_norm": 0.34668582677841187,
+      "learning_rate": 1.6040009431145266e-05,
+      "loss": 0.1088,
+      "step": 15530
+    },
+    {
+      "epoch": 406.27450980392155,
+      "grad_norm": 0.4414779841899872,
+      "learning_rate": 1.5950306886876366e-05,
+      "loss": 0.1117,
+      "step": 15540
+    },
+    {
+      "epoch": 406.5359477124183,
+      "grad_norm": 0.3160878121852875,
+      "learning_rate": 1.586083413125906e-05,
+      "loss": 0.1113,
+      "step": 15550
+    },
+    {
+      "epoch": 406.797385620915,
+      "grad_norm": 0.3153967261314392,
+      "learning_rate": 1.577159140890835e-05,
+      "loss": 0.1136,
+      "step": 15560
+    },
+    {
+      "epoch": 407.05882352941177,
+      "grad_norm": 0.29611125588417053,
+      "learning_rate": 1.568257896381049e-05,
+      "loss": 0.1119,
+      "step": 15570
+    },
+    {
+      "epoch": 407.32026143790847,
+      "grad_norm": 0.3124426603317261,
+      "learning_rate": 1.5593797039322076e-05,
+      "loss": 0.1136,
+      "step": 15580
+    },
+    {
+      "epoch": 407.58169934640523,
+      "grad_norm": 0.29556897282600403,
+      "learning_rate": 1.5505245878169528e-05,
+      "loss": 0.1133,
+      "step": 15590
+    },
+    {
+      "epoch": 407.84313725490193,
+      "grad_norm": 0.3371742069721222,
+      "learning_rate": 1.541692572244833e-05,
+      "loss": 0.1084,
+      "step": 15600
+    },
+    {
+      "epoch": 408.1045751633987,
+      "grad_norm": 0.3488721549510956,
+      "learning_rate": 1.5328836813622393e-05,
+      "loss": 0.1134,
+      "step": 15610
+    },
+    {
+      "epoch": 408.36601307189545,
+      "grad_norm": 0.31725990772247314,
+      "learning_rate": 1.5240979392523458e-05,
+      "loss": 0.113,
+      "step": 15620
+    },
+    {
+      "epoch": 408.62745098039215,
+      "grad_norm": 0.37460392713546753,
+      "learning_rate": 1.5153353699350337e-05,
+      "loss": 0.1115,
+      "step": 15630
+    },
+    {
+      "epoch": 408.8888888888889,
+      "grad_norm": 0.3097977936267853,
+      "learning_rate": 1.5065959973668353e-05,
+      "loss": 0.1125,
+      "step": 15640
+    },
+    {
+      "epoch": 409.1503267973856,
+      "grad_norm": 0.27604126930236816,
+      "learning_rate": 1.4978798454408605e-05,
+      "loss": 0.1116,
+      "step": 15650
+    },
+    {
+      "epoch": 409.4117647058824,
+      "grad_norm": 0.30955255031585693,
+      "learning_rate": 1.489186937986734e-05,
+      "loss": 0.11,
+      "step": 15660
+    },
+    {
+      "epoch": 409.6732026143791,
+      "grad_norm": 0.310028076171875,
+      "learning_rate": 1.4805172987705362e-05,
+      "loss": 0.1135,
+      "step": 15670
+    },
+    {
+      "epoch": 409.93464052287584,
+      "grad_norm": 0.3490021824836731,
+      "learning_rate": 1.471870951494726e-05,
+      "loss": 0.1142,
+      "step": 15680
+    },
+    {
+      "epoch": 410.19607843137254,
+      "grad_norm": 0.40562987327575684,
+      "learning_rate": 1.4641091730943024e-05,
+      "loss": 0.1104,
+      "step": 15690
+    },
+    {
+      "epoch": 410.4575163398693,
+      "grad_norm": 0.33675485849380493,
+      "learning_rate": 1.4555071455773993e-05,
+      "loss": 0.1106,
+      "step": 15700
+    },
+    {
+      "epoch": 410.718954248366,
+      "grad_norm": 0.3096622824668884,
+      "learning_rate": 1.4469284783776893e-05,
+      "loss": 0.1133,
+      "step": 15710
+    },
+    {
+      "epoch": 410.98039215686276,
+      "grad_norm": 0.3052927851676941,
+      "learning_rate": 1.43837319494892e-05,
+      "loss": 0.1142,
+      "step": 15720
+    },
+    {
+      "epoch": 411.24183006535947,
+      "grad_norm": 0.3547520935535431,
+      "learning_rate": 1.4298413186809123e-05,
+      "loss": 0.1137,
+      "step": 15730
+    },
+    {
+      "epoch": 411.5032679738562,
+      "grad_norm": 0.3303002417087555,
+      "learning_rate": 1.4213328728994857e-05,
+      "loss": 0.1126,
+      "step": 15740
+    },
+    {
+      "epoch": 411.7647058823529,
+      "grad_norm": 0.3191388249397278,
+      "learning_rate": 1.4128478808664125e-05,
+      "loss": 0.1101,
+      "step": 15750
+    },
+    {
+      "epoch": 412.0261437908497,
+      "grad_norm": 0.29005515575408936,
+      "learning_rate": 1.4043863657793332e-05,
+      "loss": 0.1122,
+      "step": 15760
+    },
+    {
+      "epoch": 412.2875816993464,
+      "grad_norm": 0.33250725269317627,
+      "learning_rate": 1.3959483507717042e-05,
+      "loss": 0.1104,
+      "step": 15770
+    },
+    {
+      "epoch": 412.54901960784315,
+      "grad_norm": 0.2780681252479553,
+      "learning_rate": 1.3875338589127418e-05,
+      "loss": 0.1114,
+      "step": 15780
+    },
+    {
+      "epoch": 412.81045751633985,
+      "grad_norm": 0.35173699259757996,
+      "learning_rate": 1.3791429132073408e-05,
+      "loss": 0.1113,
+      "step": 15790
+    },
+    {
+      "epoch": 413.0718954248366,
+      "grad_norm": 0.3443647027015686,
+      "learning_rate": 1.3707755365960317e-05,
+      "loss": 0.1156,
+      "step": 15800
+    },
+    {
+      "epoch": 413.3333333333333,
+      "grad_norm": 0.2871156930923462,
+      "learning_rate": 1.3624317519548979e-05,
+      "loss": 0.1131,
+      "step": 15810
+    },
+    {
+      "epoch": 413.5947712418301,
+      "grad_norm": 0.26214319467544556,
+      "learning_rate": 1.3541115820955285e-05,
+      "loss": 0.1083,
+      "step": 15820
+    },
+    {
+      "epoch": 413.8562091503268,
+      "grad_norm": 0.3014472723007202,
+      "learning_rate": 1.3458150497649525e-05,
+      "loss": 0.1142,
+      "step": 15830
+    },
+    {
+      "epoch": 414.11764705882354,
+      "grad_norm": 0.3216531276702881,
+      "learning_rate": 1.3375421776455699e-05,
+      "loss": 0.1124,
+      "step": 15840
+    },
+    {
+      "epoch": 414.37908496732024,
+      "grad_norm": 0.3679593503475189,
+      "learning_rate": 1.3292929883550998e-05,
+      "loss": 0.1134,
+      "step": 15850
+    },
+    {
+      "epoch": 414.640522875817,
+      "grad_norm": 0.2999366819858551,
+      "learning_rate": 1.3210675044465103e-05,
+      "loss": 0.1124,
+      "step": 15860
+    },
+    {
+      "epoch": 414.9019607843137,
+      "grad_norm": 0.34971389174461365,
+      "learning_rate": 1.3128657484079566e-05,
+      "loss": 0.1131,
+      "step": 15870
+    },
+    {
+      "epoch": 415.16339869281046,
+      "grad_norm": 0.3751276433467865,
+      "learning_rate": 1.3046877426627313e-05,
+      "loss": 0.1118,
+      "step": 15880
+    },
+    {
+      "epoch": 415.42483660130716,
+      "grad_norm": 0.32183781266212463,
+      "learning_rate": 1.2965335095691889e-05,
+      "loss": 0.1129,
+      "step": 15890
+    },
+    {
+      "epoch": 415.6862745098039,
+      "grad_norm": 0.35736966133117676,
+      "learning_rate": 1.2884030714206874e-05,
+      "loss": 0.1109,
+      "step": 15900
+    },
+    {
+      "epoch": 415.9477124183006,
+      "grad_norm": 0.3015005588531494,
+      "learning_rate": 1.2802964504455395e-05,
+      "loss": 0.1124,
+      "step": 15910
+    },
+    {
+      "epoch": 416.2091503267974,
+      "grad_norm": 0.34793102741241455,
+      "learning_rate": 1.272213668806933e-05,
+      "loss": 0.1146,
+      "step": 15920
+    },
+    {
+      "epoch": 416.47058823529414,
+      "grad_norm": 0.3514239490032196,
+      "learning_rate": 1.2641547486028882e-05,
+      "loss": 0.1106,
+      "step": 15930
+    },
+    {
+      "epoch": 416.73202614379085,
+      "grad_norm": 0.31007125973701477,
+      "learning_rate": 1.2561197118661828e-05,
+      "loss": 0.1102,
+      "step": 15940
+    },
+    {
+      "epoch": 416.9934640522876,
+      "grad_norm": 0.31926625967025757,
+      "learning_rate": 1.2481085805643e-05,
+      "loss": 0.1129,
+      "step": 15950
+    },
+    {
+      "epoch": 417.2549019607843,
+      "grad_norm": 0.3109895884990692,
+      "learning_rate": 1.2401213765993691e-05,
+      "loss": 0.1124,
+      "step": 15960
+    },
+    {
+      "epoch": 417.51633986928107,
+      "grad_norm": 0.33066773414611816,
+      "learning_rate": 1.2321581218080979e-05,
+      "loss": 0.112,
+      "step": 15970
+    },
+    {
+      "epoch": 417.77777777777777,
+      "grad_norm": 0.4178493618965149,
+      "learning_rate": 1.2242188379617236e-05,
+      "loss": 0.1121,
+      "step": 15980
+    },
+    {
+      "epoch": 418.03921568627453,
+      "grad_norm": 0.3649527430534363,
+      "learning_rate": 1.2163035467659444e-05,
+      "loss": 0.1127,
+      "step": 15990
+    },
+    {
+      "epoch": 418.30065359477123,
+      "grad_norm": 0.29703032970428467,
+      "learning_rate": 1.2084122698608625e-05,
+      "loss": 0.1118,
+      "step": 16000
+    },
+    {
+      "epoch": 418.562091503268,
+      "grad_norm": 0.31829434633255005,
+      "learning_rate": 1.2005450288209297e-05,
+      "loss": 0.1123,
+      "step": 16010
+    },
+    {
+      "epoch": 418.8235294117647,
+      "grad_norm": 0.3175608217716217,
+      "learning_rate": 1.1927018451548811e-05,
+      "loss": 0.1096,
+      "step": 16020
+    },
+    {
+      "epoch": 419.08496732026146,
+      "grad_norm": 0.30831390619277954,
+      "learning_rate": 1.1848827403056828e-05,
+      "loss": 0.1153,
+      "step": 16030
+    },
+    {
+      "epoch": 419.34640522875816,
+      "grad_norm": 0.3015325665473938,
+      "learning_rate": 1.1770877356504683e-05,
+      "loss": 0.1087,
+      "step": 16040
+    },
+    {
+      "epoch": 419.6078431372549,
+      "grad_norm": 0.362508624792099,
+      "learning_rate": 1.1693168525004805e-05,
+      "loss": 0.113,
+      "step": 16050
+    },
+    {
+      "epoch": 419.8692810457516,
+      "grad_norm": 0.36422449350357056,
+      "learning_rate": 1.1615701121010214e-05,
+      "loss": 0.1133,
+      "step": 16060
+    },
+    {
+      "epoch": 420.1307189542484,
+      "grad_norm": 0.27966681122779846,
+      "learning_rate": 1.1538475356313794e-05,
+      "loss": 0.1118,
+      "step": 16070
+    },
+    {
+      "epoch": 420.3921568627451,
+      "grad_norm": 0.362490713596344,
+      "learning_rate": 1.1461491442047878e-05,
+      "loss": 0.1122,
+      "step": 16080
+    },
+    {
+      "epoch": 420.65359477124184,
+      "grad_norm": 0.3337494432926178,
+      "learning_rate": 1.138474958868352e-05,
+      "loss": 0.11,
+      "step": 16090
+    },
+    {
+      "epoch": 420.91503267973854,
+      "grad_norm": 0.3513115346431732,
+      "learning_rate": 1.1308250006029997e-05,
+      "loss": 0.1138,
+      "step": 16100
+    },
+    {
+      "epoch": 421.1764705882353,
+      "grad_norm": 0.28625595569610596,
+      "learning_rate": 1.123199290323429e-05,
+      "loss": 0.1115,
+      "step": 16110
+    },
+    {
+      "epoch": 421.437908496732,
+      "grad_norm": 0.3301337957382202,
+      "learning_rate": 1.1155978488780384e-05,
+      "loss": 0.1113,
+      "step": 16120
+    },
+    {
+      "epoch": 421.69934640522877,
+      "grad_norm": 0.42644599080085754,
+      "learning_rate": 1.1080206970488793e-05,
+      "loss": 0.1157,
+      "step": 16130
+    },
+    {
+      "epoch": 421.96078431372547,
+      "grad_norm": 0.35698893666267395,
+      "learning_rate": 1.1004678555515957e-05,
+      "loss": 0.1112,
+      "step": 16140
+    },
+    {
+      "epoch": 422.22222222222223,
+      "grad_norm": 0.37550118565559387,
+      "learning_rate": 1.0929393450353654e-05,
+      "loss": 0.1127,
+      "step": 16150
+    },
+    {
+      "epoch": 422.48366013071893,
+      "grad_norm": 0.31585901975631714,
+      "learning_rate": 1.0854351860828527e-05,
+      "loss": 0.1109,
+      "step": 16160
+    },
+    {
+      "epoch": 422.7450980392157,
+      "grad_norm": 0.3062879741191864,
+      "learning_rate": 1.0779553992101387e-05,
+      "loss": 0.1123,
+      "step": 16170
+    },
+    {
+      "epoch": 423.0065359477124,
+      "grad_norm": 0.38038599491119385,
+      "learning_rate": 1.0705000048666735e-05,
+      "loss": 0.1117,
+      "step": 16180
+    },
+    {
+      "epoch": 423.26797385620915,
+      "grad_norm": 0.30192792415618896,
+      "learning_rate": 1.0630690234352259e-05,
+      "loss": 0.1096,
+      "step": 16190
+    },
+    {
+      "epoch": 423.52941176470586,
+      "grad_norm": 0.2793295383453369,
+      "learning_rate": 1.0556624752318101e-05,
+      "loss": 0.1139,
+      "step": 16200
+    },
+    {
+      "epoch": 423.7908496732026,
+      "grad_norm": 0.35322046279907227,
+      "learning_rate": 1.0482803805056507e-05,
+      "loss": 0.1133,
+      "step": 16210
+    },
+    {
+      "epoch": 424.0522875816994,
+      "grad_norm": 0.33232808113098145,
+      "learning_rate": 1.0409227594391102e-05,
+      "loss": 0.1096,
+      "step": 16220
+    },
+    {
+      "epoch": 424.3137254901961,
+      "grad_norm": 0.2976301908493042,
+      "learning_rate": 1.0335896321476413e-05,
+      "loss": 0.1115,
+      "step": 16230
+    },
+    {
+      "epoch": 424.57516339869284,
+      "grad_norm": 0.35135358572006226,
+      "learning_rate": 1.0262810186797389e-05,
+      "loss": 0.1116,
+      "step": 16240
+    },
+    {
+      "epoch": 424.83660130718954,
+      "grad_norm": 0.31133124232292175,
+      "learning_rate": 1.0189969390168696e-05,
+      "loss": 0.1139,
+      "step": 16250
+    },
+    {
+      "epoch": 425.0980392156863,
+      "grad_norm": 0.3206573724746704,
+      "learning_rate": 1.0117374130734314e-05,
+      "loss": 0.1106,
+      "step": 16260
+    },
+    {
+      "epoch": 425.359477124183,
+      "grad_norm": 0.29747602343559265,
+      "learning_rate": 1.0045024606966902e-05,
+      "loss": 0.1112,
+      "step": 16270
+    },
+    {
+      "epoch": 425.62091503267976,
+      "grad_norm": 0.3697657585144043,
+      "learning_rate": 9.972921016667269e-06,
+      "loss": 0.1128,
+      "step": 16280
+    },
+    {
+      "epoch": 425.88235294117646,
+      "grad_norm": 0.3043394088745117,
+      "learning_rate": 9.90106355696393e-06,
+      "loss": 0.1128,
+      "step": 16290
+    },
+    {
+      "epoch": 426.1437908496732,
+      "grad_norm": 0.34919267892837524,
+      "learning_rate": 9.82945242431238e-06,
+      "loss": 0.111,
+      "step": 16300
+    },
+    {
+      "epoch": 426.4052287581699,
+      "grad_norm": 0.3754013478755951,
+      "learning_rate": 9.758087814494764e-06,
+      "loss": 0.1133,
+      "step": 16310
+    },
+    {
+      "epoch": 426.6666666666667,
+      "grad_norm": 0.2977403402328491,
+      "learning_rate": 9.686969922619193e-06,
+      "loss": 0.1112,
+      "step": 16320
+    },
+    {
+      "epoch": 426.9281045751634,
+      "grad_norm": 0.27013343572616577,
+      "learning_rate": 9.616098943119234e-06,
+      "loss": 0.1104,
+      "step": 16330
+    },
+    {
+      "epoch": 427.18954248366015,
+      "grad_norm": 0.35326704382896423,
+      "learning_rate": 9.545475069753484e-06,
+      "loss": 0.1113,
+      "step": 16340
+    },
+    {
+      "epoch": 427.45098039215685,
+      "grad_norm": 0.39166802167892456,
+      "learning_rate": 9.475098495604884e-06,
+      "loss": 0.1082,
+      "step": 16350
+    },
+    {
+      "epoch": 427.7124183006536,
+      "grad_norm": 0.3371388614177704,
+      "learning_rate": 9.404969413080322e-06,
+      "loss": 0.1146,
+      "step": 16360
+    },
+    {
+      "epoch": 427.9738562091503,
+      "grad_norm": 0.3773285448551178,
+      "learning_rate": 9.335088013910021e-06,
+      "loss": 0.116,
+      "step": 16370
+    },
+    {
+      "epoch": 428.2352941176471,
+      "grad_norm": 0.31114351749420166,
+      "learning_rate": 9.265454489147052e-06,
+      "loss": 0.1115,
+      "step": 16380
+    },
+    {
+      "epoch": 428.4967320261438,
+      "grad_norm": 0.31893211603164673,
+      "learning_rate": 9.196069029166831e-06,
+      "loss": 0.1146,
+      "step": 16390
+    },
+    {
+      "epoch": 428.75816993464053,
+      "grad_norm": 0.3122820556163788,
+      "learning_rate": 9.126931823666517e-06,
+      "loss": 0.1105,
+      "step": 16400
+    },
+    {
+      "epoch": 429.01960784313724,
+      "grad_norm": 0.41064098477363586,
+      "learning_rate": 9.058043061664655e-06,
+      "loss": 0.1125,
+      "step": 16410
+    },
+    {
+      "epoch": 429.281045751634,
+      "grad_norm": 0.30506137013435364,
+      "learning_rate": 8.989402931500434e-06,
+      "loss": 0.1094,
+      "step": 16420
+    },
+    {
+      "epoch": 429.5424836601307,
+      "grad_norm": 0.298270583152771,
+      "learning_rate": 8.921011620833364e-06,
+      "loss": 0.1134,
+      "step": 16430
+    },
+    {
+      "epoch": 429.80392156862746,
+      "grad_norm": 0.3039408326148987,
+      "learning_rate": 8.852869316642688e-06,
+      "loss": 0.1129,
+      "step": 16440
+    },
+    {
+      "epoch": 430.06535947712416,
+      "grad_norm": 0.33438700437545776,
+      "learning_rate": 8.78497620522687e-06,
+      "loss": 0.1124,
+      "step": 16450
+    },
+    {
+      "epoch": 430.3267973856209,
+      "grad_norm": 0.3313377797603607,
+      "learning_rate": 8.717332472203033e-06,
+      "loss": 0.1095,
+      "step": 16460
+    },
+    {
+      "epoch": 430.5882352941176,
+      "grad_norm": 0.3241206705570221,
+      "learning_rate": 8.649938302506633e-06,
+      "loss": 0.1133,
+      "step": 16470
+    },
+    {
+      "epoch": 430.8496732026144,
+      "grad_norm": 0.3204852342605591,
+      "learning_rate": 8.582793880390693e-06,
+      "loss": 0.1111,
+      "step": 16480
+    },
+    {
+      "epoch": 431.1111111111111,
+      "grad_norm": 0.315046489238739,
+      "learning_rate": 8.515899389425542e-06,
+      "loss": 0.1135,
+      "step": 16490
+    },
+    {
+      "epoch": 431.37254901960785,
+      "grad_norm": 0.3156817555427551,
+      "learning_rate": 8.449255012498148e-06,
+      "loss": 0.1084,
+      "step": 16500
+    },
+    {
+      "epoch": 431.63398692810455,
+      "grad_norm": 0.33890920877456665,
+      "learning_rate": 8.382860931811687e-06,
+      "loss": 0.113,
+      "step": 16510
+    },
+    {
+      "epoch": 431.8954248366013,
+      "grad_norm": 0.3318261206150055,
+      "learning_rate": 8.31671732888506e-06,
+      "loss": 0.1128,
+      "step": 16520
+    },
+    {
+      "epoch": 432.15686274509807,
+      "grad_norm": 0.3495106101036072,
+      "learning_rate": 8.250824384552314e-06,
+      "loss": 0.1134,
+      "step": 16530
+    },
+    {
+      "epoch": 432.41830065359477,
+      "grad_norm": 0.3283674120903015,
+      "learning_rate": 8.185182278962288e-06,
+      "loss": 0.112,
+      "step": 16540
+    },
+    {
+      "epoch": 432.67973856209153,
+      "grad_norm": 0.3081677556037903,
+      "learning_rate": 8.119791191577975e-06,
+      "loss": 0.112,
+      "step": 16550
+    },
+    {
+      "epoch": 432.94117647058823,
+      "grad_norm": 0.3371969759464264,
+      "learning_rate": 8.054651301176087e-06,
+      "loss": 0.1108,
+      "step": 16560
+    },
+    {
+      "epoch": 433.202614379085,
+      "grad_norm": 0.3539562225341797,
+      "learning_rate": 7.989762785846633e-06,
+      "loss": 0.1116,
+      "step": 16570
+    },
+    {
+      "epoch": 433.4640522875817,
+      "grad_norm": 0.36196842789649963,
+      "learning_rate": 7.925125822992307e-06,
+      "loss": 0.1129,
+      "step": 16580
+    },
+    {
+      "epoch": 433.72549019607845,
+      "grad_norm": 0.3191837966442108,
+      "learning_rate": 7.860740589328142e-06,
+      "loss": 0.1118,
+      "step": 16590
+    },
+    {
+      "epoch": 433.98692810457516,
+      "grad_norm": 0.31484946608543396,
+      "learning_rate": 7.796607260880839e-06,
+      "loss": 0.1124,
+      "step": 16600
+    },
+    {
+      "epoch": 434.2483660130719,
+      "grad_norm": 0.33817651867866516,
+      "learning_rate": 7.73272601298851e-06,
+      "loss": 0.1107,
+      "step": 16610
+    },
+    {
+      "epoch": 434.5098039215686,
+      "grad_norm": 0.3488980531692505,
+      "learning_rate": 7.669097020300064e-06,
+      "loss": 0.1137,
+      "step": 16620
+    },
+    {
+      "epoch": 434.7712418300654,
+      "grad_norm": 0.378192275762558,
+      "learning_rate": 7.605720456774701e-06,
+      "loss": 0.1099,
+      "step": 16630
+    },
+    {
+      "epoch": 435.0326797385621,
+      "grad_norm": 0.4347194731235504,
+      "learning_rate": 7.542596495681575e-06,
+      "loss": 0.1138,
+      "step": 16640
+    },
+    {
+      "epoch": 435.29411764705884,
+      "grad_norm": 0.314876526594162,
+      "learning_rate": 7.479725309599117e-06,
+      "loss": 0.1113,
+      "step": 16650
+    },
+    {
+      "epoch": 435.55555555555554,
+      "grad_norm": 0.3719530999660492,
+      "learning_rate": 7.417107070414786e-06,
+      "loss": 0.1121,
+      "step": 16660
+    },
+    {
+      "epoch": 435.8169934640523,
+      "grad_norm": 0.31676945090293884,
+      "learning_rate": 7.354741949324473e-06,
+      "loss": 0.1122,
+      "step": 16670
+    },
+    {
+      "epoch": 436.078431372549,
+      "grad_norm": 0.3606952726840973,
+      "learning_rate": 7.292630116832011e-06,
+      "loss": 0.1109,
+      "step": 16680
+    },
+    {
+      "epoch": 436.33986928104576,
+      "grad_norm": 0.36602726578712463,
+      "learning_rate": 7.23077174274881e-06,
+      "loss": 0.1135,
+      "step": 16690
+    },
+    {
+      "epoch": 436.60130718954247,
+      "grad_norm": 0.359847754240036,
+      "learning_rate": 7.169166996193255e-06,
+      "loss": 0.108,
+      "step": 16700
+    },
+    {
+      "epoch": 436.8627450980392,
+      "grad_norm": 0.3841437101364136,
+      "learning_rate": 7.1078160455903875e-06,
+      "loss": 0.1148,
+      "step": 16710
+    },
+    {
+      "epoch": 437.12418300653593,
+      "grad_norm": 0.3049101233482361,
+      "learning_rate": 7.0467190586713915e-06,
+      "loss": 0.11,
+      "step": 16720
+    },
+    {
+      "epoch": 437.3856209150327,
+      "grad_norm": 0.32839474081993103,
+      "learning_rate": 6.985876202473085e-06,
+      "loss": 0.1117,
+      "step": 16730
+    },
+    {
+      "epoch": 437.6470588235294,
+      "grad_norm": 0.3973640501499176,
+      "learning_rate": 6.925287643337497e-06,
+      "loss": 0.1119,
+      "step": 16740
+    },
+    {
+      "epoch": 437.90849673202615,
+      "grad_norm": 0.404038667678833,
+      "learning_rate": 6.864953546911424e-06,
+      "loss": 0.1126,
+      "step": 16750
+    },
+    {
+      "epoch": 438.16993464052285,
+      "grad_norm": 0.44833749532699585,
+      "learning_rate": 6.8048740781460065e-06,
+      "loss": 0.1101,
+      "step": 16760
+    },
+    {
+      "epoch": 438.4313725490196,
+      "grad_norm": 0.35082849860191345,
+      "learning_rate": 6.7450494012962326e-06,
+      "loss": 0.1105,
+      "step": 16770
+    },
+    {
+      "epoch": 438.6928104575163,
+      "grad_norm": 0.3201982080936432,
+      "learning_rate": 6.685479679920459e-06,
+      "loss": 0.1132,
+      "step": 16780
+    },
+    {
+      "epoch": 438.9542483660131,
+      "grad_norm": 0.3554164171218872,
+      "learning_rate": 6.626165076880031e-06,
+      "loss": 0.1115,
+      "step": 16790
+    },
+    {
+      "epoch": 439.2156862745098,
+      "grad_norm": 0.4029858410358429,
+      "learning_rate": 6.5671057543387985e-06,
+      "loss": 0.1135,
+      "step": 16800
+    },
+    {
+      "epoch": 439.47712418300654,
+      "grad_norm": 0.3851124346256256,
+      "learning_rate": 6.508301873762712e-06,
+      "loss": 0.1112,
+      "step": 16810
+    },
+    {
+      "epoch": 439.73856209150324,
+      "grad_norm": 0.2964167594909668,
+      "learning_rate": 6.449753595919361e-06,
+      "loss": 0.1131,
+      "step": 16820
+    },
+    {
+      "epoch": 440.0,
+      "grad_norm": 0.3902610242366791,
+      "learning_rate": 6.391461080877436e-06,
+      "loss": 0.1112,
+      "step": 16830
+    },
+    {
+      "epoch": 440.26143790849676,
+      "grad_norm": 0.3522237241268158,
+      "learning_rate": 6.333424488006501e-06,
+      "loss": 0.1104,
+      "step": 16840
+    },
+    {
+      "epoch": 440.52287581699346,
+      "grad_norm": 0.3239668309688568,
+      "learning_rate": 6.275643975976353e-06,
+      "loss": 0.1129,
+      "step": 16850
+    },
+    {
+      "epoch": 440.7843137254902,
+      "grad_norm": 0.3547097444534302,
+      "learning_rate": 6.218119702756708e-06,
+      "loss": 0.1116,
+      "step": 16860
+    },
+    {
+      "epoch": 441.0457516339869,
+      "grad_norm": 0.29542434215545654,
+      "learning_rate": 6.160851825616787e-06,
+      "loss": 0.1101,
+      "step": 16870
+    },
+    {
+      "epoch": 441.3071895424837,
+      "grad_norm": 0.3131840229034424,
+      "learning_rate": 6.103840501124702e-06,
+      "loss": 0.1116,
+      "step": 16880
+    },
+    {
+      "epoch": 441.5686274509804,
+      "grad_norm": 0.32945704460144043,
+      "learning_rate": 6.047085885147286e-06,
+      "loss": 0.1125,
+      "step": 16890
+    },
+    {
+      "epoch": 441.83006535947715,
+      "grad_norm": 0.3166607618331909,
+      "learning_rate": 5.990588132849462e-06,
+      "loss": 0.1084,
+      "step": 16900
+    },
+    {
+      "epoch": 442.09150326797385,
+      "grad_norm": 0.4009234607219696,
+      "learning_rate": 5.9343473986939405e-06,
+      "loss": 0.1174,
+      "step": 16910
+    },
+    {
+      "epoch": 442.3529411764706,
+      "grad_norm": 0.3329063653945923,
+      "learning_rate": 5.87836383644077e-06,
+      "loss": 0.1099,
+      "step": 16920
+    },
+    {
+      "epoch": 442.6143790849673,
+      "grad_norm": 0.3919806480407715,
+      "learning_rate": 5.8226375991468294e-06,
+      "loss": 0.1101,
+      "step": 16930
+    },
+    {
+      "epoch": 442.87581699346407,
+      "grad_norm": 0.35498499870300293,
+      "learning_rate": 5.767168839165538e-06,
+      "loss": 0.1122,
+      "step": 16940
+    },
+    {
+      "epoch": 443.1372549019608,
+      "grad_norm": 0.3532790243625641,
+      "learning_rate": 5.711957708146365e-06,
+      "loss": 0.1116,
+      "step": 16950
+    },
+    {
+      "epoch": 443.39869281045753,
+      "grad_norm": 0.2659851014614105,
+      "learning_rate": 5.657004357034445e-06,
+      "loss": 0.1128,
+      "step": 16960
+    },
+    {
+      "epoch": 443.66013071895424,
+      "grad_norm": 0.3570290207862854,
+      "learning_rate": 5.602308936070133e-06,
+      "loss": 0.1143,
+      "step": 16970
+    },
+    {
+      "epoch": 443.921568627451,
+      "grad_norm": 0.30066853761672974,
+      "learning_rate": 5.547871594788611e-06,
+      "loss": 0.1106,
+      "step": 16980
+    },
+    {
+      "epoch": 444.1830065359477,
+      "grad_norm": 0.30849123001098633,
+      "learning_rate": 5.49369248201953e-06,
+      "loss": 0.1101,
+      "step": 16990
+    },
+    {
+      "epoch": 444.44444444444446,
+      "grad_norm": 0.30579501390457153,
+      "learning_rate": 5.4397717458864576e-06,
+      "loss": 0.1113,
+      "step": 17000
+    },
+    {
+      "epoch": 444.70588235294116,
+      "grad_norm": 0.30909931659698486,
+      "learning_rate": 5.3861095338066826e-06,
+      "loss": 0.1124,
+      "step": 17010
+    },
+    {
+      "epoch": 444.9673202614379,
+      "grad_norm": 0.3454172909259796,
+      "learning_rate": 5.332705992490616e-06,
+      "loss": 0.1124,
+      "step": 17020
+    },
+    {
+      "epoch": 445.2287581699346,
+      "grad_norm": 0.390906423330307,
+      "learning_rate": 5.279561267941491e-06,
+      "loss": 0.1097,
+      "step": 17030
+    },
+    {
+      "epoch": 445.4901960784314,
+      "grad_norm": 0.3330720365047455,
+      "learning_rate": 5.226675505454981e-06,
+      "loss": 0.1113,
+      "step": 17040
+    },
+    {
+      "epoch": 445.7516339869281,
+      "grad_norm": 0.33850574493408203,
+      "learning_rate": 5.174048849618718e-06,
+      "loss": 0.1118,
+      "step": 17050
+    },
+    {
+      "epoch": 446.01307189542484,
+      "grad_norm": 0.3519850969314575,
+      "learning_rate": 5.121681444311987e-06,
+      "loss": 0.1147,
+      "step": 17060
+    },
+    {
+      "epoch": 446.27450980392155,
+      "grad_norm": 0.3401840031147003,
+      "learning_rate": 5.069573432705277e-06,
+      "loss": 0.1116,
+      "step": 17070
+    },
+    {
+      "epoch": 446.5359477124183,
+      "grad_norm": 0.3529043197631836,
+      "learning_rate": 5.017724957259873e-06,
+      "loss": 0.1135,
+      "step": 17080
+    },
+    {
+      "epoch": 446.797385620915,
+      "grad_norm": 0.3206219971179962,
+      "learning_rate": 4.966136159727563e-06,
+      "loss": 0.1118,
+      "step": 17090
+    },
+    {
+      "epoch": 447.05882352941177,
+      "grad_norm": 0.3134680986404419,
+      "learning_rate": 4.914807181150139e-06,
+      "loss": 0.1092,
+      "step": 17100
+    },
+    {
+      "epoch": 447.32026143790847,
+      "grad_norm": 0.3726840019226074,
+      "learning_rate": 4.863738161859044e-06,
+      "loss": 0.1116,
+      "step": 17110
+    },
+    {
+      "epoch": 447.58169934640523,
+      "grad_norm": 0.41007351875305176,
+      "learning_rate": 4.812929241475062e-06,
+      "loss": 0.1136,
+      "step": 17120
+    },
+    {
+      "epoch": 447.84313725490193,
+      "grad_norm": 0.2826221287250519,
+      "learning_rate": 4.762380558907798e-06,
+      "loss": 0.1113,
+      "step": 17130
+    },
+    {
+      "epoch": 448.1045751633987,
+      "grad_norm": 0.3534366488456726,
+      "learning_rate": 4.712092252355471e-06,
+      "loss": 0.1109,
+      "step": 17140
+    },
+    {
+      "epoch": 448.36601307189545,
+      "grad_norm": 0.28506147861480713,
+      "learning_rate": 4.662064459304372e-06,
+      "loss": 0.1134,
+      "step": 17150
+    },
+    {
+      "epoch": 448.62745098039215,
+      "grad_norm": 0.3237999975681305,
+      "learning_rate": 4.612297316528547e-06,
+      "loss": 0.1113,
+      "step": 17160
+    },
+    {
+      "epoch": 448.8888888888889,
+      "grad_norm": 0.37305569648742676,
+      "learning_rate": 4.5627909600895026e-06,
+      "loss": 0.1105,
+      "step": 17170
+    },
+    {
+      "epoch": 449.1503267973856,
+      "grad_norm": 0.30823808908462524,
+      "learning_rate": 4.513545525335705e-06,
+      "loss": 0.1123,
+      "step": 17180
+    },
+    {
+      "epoch": 449.4117647058824,
+      "grad_norm": 0.3125160336494446,
+      "learning_rate": 4.464561146902302e-06,
+      "loss": 0.1104,
+      "step": 17190
+    },
+    {
+      "epoch": 449.6732026143791,
+      "grad_norm": 0.32170015573501587,
+      "learning_rate": 4.4158379587107335e-06,
+      "loss": 0.1118,
+      "step": 17200
+    },
+    {
+      "epoch": 449.93464052287584,
+      "grad_norm": 0.3177231550216675,
+      "learning_rate": 4.367376093968278e-06,
+      "loss": 0.1134,
+      "step": 17210
+    },
+    {
+      "epoch": 450.19607843137254,
+      "grad_norm": 0.35781294107437134,
+      "learning_rate": 4.319175685167887e-06,
+      "loss": 0.1095,
+      "step": 17220
+    },
+    {
+      "epoch": 450.4575163398693,
+      "grad_norm": 0.3756282925605774,
+      "learning_rate": 4.2712368640875914e-06,
+      "loss": 0.1135,
+      "step": 17230
+    },
+    {
+      "epoch": 450.718954248366,
+      "grad_norm": 0.3352264165878296,
+      "learning_rate": 4.22355976179033e-06,
+      "loss": 0.1113,
+      "step": 17240
+    },
+    {
+      "epoch": 450.98039215686276,
+      "grad_norm": 0.38803133368492126,
+      "learning_rate": 4.176144508623458e-06,
+      "loss": 0.1114,
+      "step": 17250
+    },
+    {
+      "epoch": 451.24183006535947,
+      "grad_norm": 0.3300243020057678,
+      "learning_rate": 4.128991234218471e-06,
+      "loss": 0.1115,
+      "step": 17260
+    },
+    {
+      "epoch": 451.5032679738562,
+      "grad_norm": 0.30495062470436096,
+      "learning_rate": 4.082100067490635e-06,
+      "loss": 0.1117,
+      "step": 17270
+    },
+    {
+      "epoch": 451.7647058823529,
+      "grad_norm": 0.3008281886577606,
+      "learning_rate": 4.03547113663858e-06,
+      "loss": 0.1138,
+      "step": 17280
+    },
+    {
+      "epoch": 452.0261437908497,
+      "grad_norm": 0.3080352246761322,
+      "learning_rate": 3.989104569144065e-06,
+      "loss": 0.1098,
+      "step": 17290
+    },
+    {
+      "epoch": 452.2875816993464,
+      "grad_norm": 0.30234184861183167,
+      "learning_rate": 3.943000491771487e-06,
+      "loss": 0.1114,
+      "step": 17300
+    },
+    {
+      "epoch": 452.54901960784315,
+      "grad_norm": 0.3571633994579315,
+      "learning_rate": 3.897159030567621e-06,
+      "loss": 0.1114,
+      "step": 17310
+    },
+    {
+      "epoch": 452.81045751633985,
+      "grad_norm": 0.3420047163963318,
+      "learning_rate": 3.8515803108613025e-06,
+      "loss": 0.1107,
+      "step": 17320
+    },
+    {
+      "epoch": 453.0718954248366,
+      "grad_norm": 0.38761913776397705,
+      "learning_rate": 3.806264457262976e-06,
+      "loss": 0.1129,
+      "step": 17330
+    },
+    {
+      "epoch": 453.3333333333333,
+      "grad_norm": 0.3283638656139374,
+      "learning_rate": 3.7612115936644932e-06,
+      "loss": 0.113,
+      "step": 17340
+    },
+    {
+      "epoch": 453.5947712418301,
+      "grad_norm": 0.38328078389167786,
+      "learning_rate": 3.716421843238649e-06,
+      "loss": 0.1128,
+      "step": 17350
+    },
+    {
+      "epoch": 453.8562091503268,
+      "grad_norm": 0.2841828763484955,
+      "learning_rate": 3.67189532843889e-06,
+      "loss": 0.1106,
+      "step": 17360
+    },
+    {
+      "epoch": 454.11764705882354,
+      "grad_norm": 0.39247363805770874,
+      "learning_rate": 3.627632170999029e-06,
+      "loss": 0.1132,
+      "step": 17370
+    },
+    {
+      "epoch": 454.37908496732024,
+      "grad_norm": 0.38083165884017944,
+      "learning_rate": 3.5836324919328536e-06,
+      "loss": 0.1143,
+      "step": 17380
+    },
+    {
+      "epoch": 454.640522875817,
+      "grad_norm": 0.3211730420589447,
+      "learning_rate": 3.5398964115337828e-06,
+      "loss": 0.1088,
+      "step": 17390
+    },
+    {
+      "epoch": 454.9019607843137,
+      "grad_norm": 0.347988486289978,
+      "learning_rate": 3.496424049374614e-06,
+      "loss": 0.1084,
+      "step": 17400
+    },
+    {
+      "epoch": 455.16339869281046,
+      "grad_norm": 0.331668496131897,
+      "learning_rate": 3.4532155243070963e-06,
+      "loss": 0.1148,
+      "step": 17410
+    },
+    {
+      "epoch": 455.42483660130716,
+      "grad_norm": 0.3547741174697876,
+      "learning_rate": 3.410270954461725e-06,
+      "loss": 0.1101,
+      "step": 17420
+    },
+    {
+      "epoch": 455.6862745098039,
+      "grad_norm": 0.3205719292163849,
+      "learning_rate": 3.3675904572472825e-06,
+      "loss": 0.1102,
+      "step": 17430
+    },
+    {
+      "epoch": 455.9477124183006,
+      "grad_norm": 0.3358950912952423,
+      "learning_rate": 3.3251741493506294e-06,
+      "loss": 0.1109,
+      "step": 17440
+    },
+    {
+      "epoch": 456.2091503267974,
+      "grad_norm": 0.3523581027984619,
+      "learning_rate": 3.2830221467363476e-06,
+      "loss": 0.1134,
+      "step": 17450
+    },
+    {
+      "epoch": 456.47058823529414,
+      "grad_norm": 0.36160987615585327,
+      "learning_rate": 3.2411345646463643e-06,
+      "loss": 0.1101,
+      "step": 17460
+    },
+    {
+      "epoch": 456.73202614379085,
+      "grad_norm": 0.2866756319999695,
+      "learning_rate": 3.1995115175997736e-06,
+      "loss": 0.1121,
+      "step": 17470
+    },
+    {
+      "epoch": 456.9934640522876,
+      "grad_norm": 0.31712275743484497,
+      "learning_rate": 3.1581531193923706e-06,
+      "loss": 0.1127,
+      "step": 17480
+    },
+    {
+      "epoch": 457.2549019607843,
+      "grad_norm": 0.3979257345199585,
+      "learning_rate": 3.1170594830964405e-06,
+      "loss": 0.1103,
+      "step": 17490
+    },
+    {
+      "epoch": 457.51633986928107,
+      "grad_norm": 0.3742763102054596,
+      "learning_rate": 3.0762307210604246e-06,
+      "loss": 0.1115,
+      "step": 17500
+    },
+    {
+      "epoch": 457.77777777777777,
+      "grad_norm": 0.3208019435405731,
+      "learning_rate": 3.0356669449085775e-06,
+      "loss": 0.1157,
+      "step": 17510
+    },
+    {
+      "epoch": 458.03921568627453,
+      "grad_norm": 0.31606632471084595,
+      "learning_rate": 2.9953682655407434e-06,
+      "loss": 0.1093,
+      "step": 17520
+    },
+    {
+      "epoch": 458.30065359477123,
+      "grad_norm": 0.36023128032684326,
+      "learning_rate": 2.9553347931319586e-06,
+      "loss": 0.114,
+      "step": 17530
+    },
+    {
+      "epoch": 458.562091503268,
+      "grad_norm": 0.39027029275894165,
+      "learning_rate": 2.9155666371321944e-06,
+      "loss": 0.1123,
+      "step": 17540
+    },
+    {
+      "epoch": 458.8235294117647,
+      "grad_norm": 0.28587546944618225,
+      "learning_rate": 2.876063906266102e-06,
+      "loss": 0.1114,
+      "step": 17550
+    },
+    {
+      "epoch": 459.08496732026146,
+      "grad_norm": 0.33159536123275757,
+      "learning_rate": 2.836826708532603e-06,
+      "loss": 0.1084,
+      "step": 17560
+    },
+    {
+      "epoch": 459.34640522875816,
+      "grad_norm": 0.37588268518447876,
+      "learning_rate": 2.7978551512047312e-06,
+      "loss": 0.1136,
+      "step": 17570
+    },
+    {
+      "epoch": 459.6078431372549,
+      "grad_norm": 0.309316486120224,
+      "learning_rate": 2.7591493408292256e-06,
+      "loss": 0.1107,
+      "step": 17580
+    },
+    {
+      "epoch": 459.8692810457516,
+      "grad_norm": 0.34601840376853943,
+      "learning_rate": 2.720709383226272e-06,
+      "loss": 0.1104,
+      "step": 17590
+    },
+    {
+      "epoch": 460.1307189542484,
+      "grad_norm": 0.28702256083488464,
+      "learning_rate": 2.682535383489282e-06,
+      "loss": 0.1126,
+      "step": 17600
+    },
+    {
+      "epoch": 460.3921568627451,
+      "grad_norm": 0.3772827088832855,
+      "learning_rate": 2.6446274459844712e-06,
+      "loss": 0.111,
+      "step": 17610
+    },
+    {
+      "epoch": 460.65359477124184,
+      "grad_norm": 0.3219127357006073,
+      "learning_rate": 2.606985674350737e-06,
+      "loss": 0.1089,
+      "step": 17620
+    },
+    {
+      "epoch": 460.91503267973854,
+      "grad_norm": 0.36690565943717957,
+      "learning_rate": 2.569610171499226e-06,
+      "loss": 0.1134,
+      "step": 17630
+    },
+    {
+      "epoch": 461.1764705882353,
+      "grad_norm": 0.3363436162471771,
+      "learning_rate": 2.5325010396131332e-06,
+      "loss": 0.1111,
+      "step": 17640
+    },
+    {
+      "epoch": 461.437908496732,
+      "grad_norm": 0.3152006268501282,
+      "learning_rate": 2.495658380147414e-06,
+      "loss": 0.111,
+      "step": 17650
+    },
+    {
+      "epoch": 461.69934640522877,
+      "grad_norm": 0.30644363164901733,
+      "learning_rate": 2.4590822938284854e-06,
+      "loss": 0.1126,
+      "step": 17660
+    },
+    {
+      "epoch": 461.96078431372547,
+      "grad_norm": 0.4446253776550293,
+      "learning_rate": 2.4227728806539672e-06,
+      "loss": 0.1135,
+      "step": 17670
+    },
+    {
+      "epoch": 462.22222222222223,
+      "grad_norm": 0.323368638753891,
+      "learning_rate": 2.386730239892432e-06,
+      "loss": 0.1088,
+      "step": 17680
+    },
+    {
+      "epoch": 462.48366013071893,
+      "grad_norm": 0.37691059708595276,
+      "learning_rate": 2.3509544700830556e-06,
+      "loss": 0.1113,
+      "step": 17690
+    },
+    {
+      "epoch": 462.7450980392157,
+      "grad_norm": 0.32642048597335815,
+      "learning_rate": 2.318984532773427e-06,
+      "loss": 0.1136,
+      "step": 17700
+    },
+    {
+      "epoch": 463.0065359477124,
+      "grad_norm": 0.37580016255378723,
+      "learning_rate": 2.283716086635357e-06,
+      "loss": 0.1124,
+      "step": 17710
+    },
+    {
+      "epoch": 463.26797385620915,
+      "grad_norm": 0.3199895918369293,
+      "learning_rate": 2.248714793086215e-06,
+      "loss": 0.112,
+      "step": 17720
+    },
+    {
+      "epoch": 463.52941176470586,
+      "grad_norm": 0.2862391471862793,
+      "learning_rate": 2.213980747818201e-06,
+      "loss": 0.111,
+      "step": 17730
+    },
+    {
+      "epoch": 463.7908496732026,
+      "grad_norm": 0.3182966411113739,
+      "learning_rate": 2.179514045792885e-06,
+      "loss": 0.111,
+      "step": 17740
+    },
+    {
+      "epoch": 464.0522875816994,
+      "grad_norm": 0.4070954918861389,
+      "learning_rate": 2.1453147812408925e-06,
+      "loss": 0.1133,
+      "step": 17750
+    },
+    {
+      "epoch": 464.3137254901961,
+      "grad_norm": 0.39886951446533203,
+      "learning_rate": 2.1113830476617193e-06,
+      "loss": 0.1129,
+      "step": 17760
+    },
+    {
+      "epoch": 464.57516339869284,
+      "grad_norm": 0.3158631920814514,
+      "learning_rate": 2.0777189378234143e-06,
+      "loss": 0.1114,
+      "step": 17770
+    },
+    {
+      "epoch": 464.83660130718954,
+      "grad_norm": 0.3350854814052582,
+      "learning_rate": 2.0443225437624e-06,
+      "loss": 0.1103,
+      "step": 17780
+    },
+    {
+      "epoch": 465.0980392156863,
+      "grad_norm": 0.3238007128238678,
+      "learning_rate": 2.0111939567831197e-06,
+      "loss": 0.11,
+      "step": 17790
+    },
+    {
+      "epoch": 465.359477124183,
+      "grad_norm": 0.3874686360359192,
+      "learning_rate": 1.9783332674578546e-06,
+      "loss": 0.1122,
+      "step": 17800
+    },
+    {
+      "epoch": 465.62091503267976,
+      "grad_norm": 0.3172845244407654,
+      "learning_rate": 1.9457405656264973e-06,
+      "loss": 0.1108,
+      "step": 17810
+    },
+    {
+      "epoch": 465.88235294117646,
+      "grad_norm": 0.3653365969657898,
+      "learning_rate": 1.913415940396246e-06,
+      "loss": 0.1124,
+      "step": 17820
+    },
+    {
+      "epoch": 466.1437908496732,
+      "grad_norm": 0.33903956413269043,
+      "learning_rate": 1.8813594801413758e-06,
+      "loss": 0.1124,
+      "step": 17830
+    },
+    {
+      "epoch": 466.4052287581699,
+      "grad_norm": 0.3369342088699341,
+      "learning_rate": 1.8495712725030478e-06,
+      "loss": 0.1157,
+      "step": 17840
+    },
+    {
+      "epoch": 466.6666666666667,
+      "grad_norm": 0.36298859119415283,
+      "learning_rate": 1.8180514043889763e-06,
+      "loss": 0.1097,
+      "step": 17850
+    },
+    {
+      "epoch": 466.9281045751634,
+      "grad_norm": 0.3302154541015625,
+      "learning_rate": 1.7867999619733179e-06,
+      "loss": 0.1104,
+      "step": 17860
+    },
+    {
+      "epoch": 467.18954248366015,
+      "grad_norm": 0.42841556668281555,
+      "learning_rate": 1.755817030696294e-06,
+      "loss": 0.1101,
+      "step": 17870
+    },
+    {
+      "epoch": 467.45098039215685,
+      "grad_norm": 0.33162766695022583,
+      "learning_rate": 1.725102695264058e-06,
+      "loss": 0.1095,
+      "step": 17880
+    },
+    {
+      "epoch": 467.7124183006536,
+      "grad_norm": 0.33728310465812683,
+      "learning_rate": 1.6946570396484507e-06,
+      "loss": 0.11,
+      "step": 17890
+    },
+    {
+      "epoch": 467.9738562091503,
+      "grad_norm": 0.4027140438556671,
+      "learning_rate": 1.6644801470867e-06,
+      "loss": 0.1145,
+      "step": 17900
+    },
+    {
+      "epoch": 468.2352941176471,
+      "grad_norm": 0.33868324756622314,
+      "learning_rate": 1.6345721000812997e-06,
+      "loss": 0.1111,
+      "step": 17910
+    },
+    {
+      "epoch": 468.4967320261438,
+      "grad_norm": 0.30771663784980774,
+      "learning_rate": 1.6049329803997092e-06,
+      "loss": 0.1131,
+      "step": 17920
+    },
+    {
+      "epoch": 468.75816993464053,
+      "grad_norm": 0.32284727692604065,
+      "learning_rate": 1.575562869074143e-06,
+      "loss": 0.1076,
+      "step": 17930
+    },
+    {
+      "epoch": 469.01960784313724,
+      "grad_norm": 0.3255426287651062,
+      "learning_rate": 1.5464618464013592e-06,
+      "loss": 0.1151,
+      "step": 17940
+    },
+    {
+      "epoch": 469.281045751634,
+      "grad_norm": 0.3239709436893463,
+      "learning_rate": 1.5176299919424486e-06,
+      "loss": 0.1113,
+      "step": 17950
+    },
+    {
+      "epoch": 469.5424836601307,
+      "grad_norm": 0.39039456844329834,
+      "learning_rate": 1.4890673845226133e-06,
+      "loss": 0.1123,
+      "step": 17960
+    },
+    {
+      "epoch": 469.80392156862746,
+      "grad_norm": 0.35394954681396484,
+      "learning_rate": 1.4607741022309106e-06,
+      "loss": 0.1157,
+      "step": 17970
+    },
+    {
+      "epoch": 470.06535947712416,
+      "grad_norm": 0.33994433283805847,
+      "learning_rate": 1.4327502224200872e-06,
+      "loss": 0.106,
+      "step": 17980
+    },
+    {
+      "epoch": 470.3267973856209,
+      "grad_norm": 0.36443567276000977,
+      "learning_rate": 1.4049958217063896e-06,
+      "loss": 0.1097,
+      "step": 17990
+    },
+    {
+      "epoch": 470.5882352941176,
+      "grad_norm": 0.28790152072906494,
+      "learning_rate": 1.3775109759692651e-06,
+      "loss": 0.1127,
+      "step": 18000
+    },
+    {
+      "epoch": 470.8496732026144,
+      "grad_norm": 0.31258511543273926,
+      "learning_rate": 1.350295760351261e-06,
+      "loss": 0.112,
+      "step": 18010
+    },
+    {
+      "epoch": 471.1111111111111,
+      "grad_norm": 0.35889947414398193,
+      "learning_rate": 1.3233502492577044e-06,
+      "loss": 0.1139,
+      "step": 18020
+    },
+    {
+      "epoch": 471.37254901960785,
+      "grad_norm": 0.31674671173095703,
+      "learning_rate": 1.2966745163566107e-06,
+      "loss": 0.1113,
+      "step": 18030
+    },
+    {
+      "epoch": 471.63398692810455,
+      "grad_norm": 0.3120609521865845,
+      "learning_rate": 1.2702686345784088e-06,
+      "loss": 0.1133,
+      "step": 18040
+    },
+    {
+      "epoch": 471.8954248366013,
+      "grad_norm": 0.4035533666610718,
+      "learning_rate": 1.2441326761157723e-06,
+      "loss": 0.1112,
+      "step": 18050
+    },
+    {
+      "epoch": 472.15686274509807,
+      "grad_norm": 0.3225405812263489,
+      "learning_rate": 1.2182667124234326e-06,
+      "loss": 0.1085,
+      "step": 18060
+    },
+    {
+      "epoch": 472.41830065359477,
+      "grad_norm": 0.3946692645549774,
+      "learning_rate": 1.1926708142179111e-06,
+      "loss": 0.1124,
+      "step": 18070
+    },
+    {
+      "epoch": 472.67973856209153,
+      "grad_norm": 0.34986528754234314,
+      "learning_rate": 1.167345051477442e-06,
+      "loss": 0.1111,
+      "step": 18080
+    },
+    {
+      "epoch": 472.94117647058823,
+      "grad_norm": 0.32098639011383057,
+      "learning_rate": 1.1422894934416838e-06,
+      "loss": 0.1115,
+      "step": 18090
+    },
+    {
+      "epoch": 473.202614379085,
+      "grad_norm": 0.33079513907432556,
+      "learning_rate": 1.117504208611586e-06,
+      "loss": 0.1116,
+      "step": 18100
+    },
+    {
+      "epoch": 473.4640522875817,
+      "grad_norm": 0.332214891910553,
+      "learning_rate": 1.092989264749167e-06,
+      "loss": 0.1106,
+      "step": 18110
+    },
+    {
+      "epoch": 473.72549019607845,
+      "grad_norm": 0.2994788885116577,
+      "learning_rate": 1.0687447288773244e-06,
+      "loss": 0.1122,
+      "step": 18120
+    },
+    {
+      "epoch": 473.98692810457516,
+      "grad_norm": 0.32751554250717163,
+      "learning_rate": 1.0447706672797264e-06,
+      "loss": 0.1126,
+      "step": 18130
+    },
+    {
+      "epoch": 474.2483660130719,
+      "grad_norm": 0.40659329295158386,
+      "learning_rate": 1.0210671455005204e-06,
+      "loss": 0.1122,
+      "step": 18140
+    },
+    {
+      "epoch": 474.5098039215686,
+      "grad_norm": 0.3787976801395416,
+      "learning_rate": 9.976342283442463e-07,
+      "loss": 0.1132,
+      "step": 18150
+    },
+    {
+      "epoch": 474.7712418300654,
+      "grad_norm": 0.32952606678009033,
+      "learning_rate": 9.744719798755907e-07,
+      "loss": 0.1134,
+      "step": 18160
+    },
+    {
+      "epoch": 475.0326797385621,
+      "grad_norm": 0.36884790658950806,
+      "learning_rate": 9.515804634192659e-07,
+      "loss": 0.1089,
+      "step": 18170
+    },
+    {
+      "epoch": 475.29411764705884,
+      "grad_norm": 0.3276387155056,
+      "learning_rate": 9.289597415597872e-07,
+      "loss": 0.1081,
+      "step": 18180
+    },
+    {
+      "epoch": 475.55555555555554,
+      "grad_norm": 0.34189870953559875,
+      "learning_rate": 9.066098761413733e-07,
+      "loss": 0.113,
+      "step": 18190
+    },
+    {
+      "epoch": 475.8169934640523,
+      "grad_norm": 0.32616567611694336,
+      "learning_rate": 8.845309282676795e-07,
+      "loss": 0.112,
+      "step": 18200
+    },
+    {
+      "epoch": 476.078431372549,
+      "grad_norm": 0.3290955126285553,
+      "learning_rate": 8.627229583017204e-07,
+      "loss": 0.1112,
+      "step": 18210
+    },
+    {
+      "epoch": 476.33986928104576,
+      "grad_norm": 0.34466055035591125,
+      "learning_rate": 8.411860258656256e-07,
+      "loss": 0.1096,
+      "step": 18220
+    },
+    {
+      "epoch": 476.60130718954247,
+      "grad_norm": 0.30991095304489136,
+      "learning_rate": 8.199201898405839e-07,
+      "loss": 0.1115,
+      "step": 18230
+    },
+    {
+      "epoch": 476.8627450980392,
+      "grad_norm": 0.34044432640075684,
+      "learning_rate": 7.989255083665659e-07,
+      "loss": 0.1145,
+      "step": 18240
+    },
+    {
+      "epoch": 477.12418300653593,
+      "grad_norm": 0.34200185537338257,
+      "learning_rate": 7.782020388422018e-07,
+      "loss": 0.1114,
+      "step": 18250
+    },
+    {
+      "epoch": 477.3856209150327,
+      "grad_norm": 0.30193594098091125,
+      "learning_rate": 7.577498379247039e-07,
+      "loss": 0.1109,
+      "step": 18260
+    },
+    {
+      "epoch": 477.6470588235294,
+      "grad_norm": 0.3074875473976135,
+      "learning_rate": 7.375689615295889e-07,
+      "loss": 0.1111,
+      "step": 18270
+    },
+    {
+      "epoch": 477.90849673202615,
+      "grad_norm": 0.3469603359699249,
+      "learning_rate": 7.176594648306111e-07,
+      "loss": 0.114,
+      "step": 18280
+    },
+    {
+      "epoch": 478.16993464052285,
+      "grad_norm": 0.2915016710758209,
+      "learning_rate": 6.980214022595744e-07,
+      "loss": 0.1096,
+      "step": 18290
+    },
+    {
+      "epoch": 478.4313725490196,
+      "grad_norm": 0.32458066940307617,
+      "learning_rate": 6.786548275061754e-07,
+      "loss": 0.1124,
+      "step": 18300
+    },
+    {
+      "epoch": 478.6928104575163,
+      "grad_norm": 0.5039242506027222,
+      "learning_rate": 6.595597935179166e-07,
+      "loss": 0.111,
+      "step": 18310
+    },
+    {
+      "epoch": 478.9542483660131,
+      "grad_norm": 0.3211175501346588,
+      "learning_rate": 6.4073635249986e-07,
+      "loss": 0.1138,
+      "step": 18320
+    },
+    {
+      "epoch": 479.2156862745098,
+      "grad_norm": 0.32088157534599304,
+      "learning_rate": 6.221845559146066e-07,
+      "loss": 0.1104,
+      "step": 18330
+    },
+    {
+      "epoch": 479.47712418300654,
+      "grad_norm": 0.38157036900520325,
+      "learning_rate": 6.039044544820404e-07,
+      "loss": 0.1124,
+      "step": 18340
+    },
+    {
+      "epoch": 479.73856209150324,
+      "grad_norm": 0.3110623061656952,
+      "learning_rate": 5.858960981792505e-07,
+      "loss": 0.1127,
+      "step": 18350
+    },
+    {
+      "epoch": 480.0,
+      "grad_norm": 0.39009323716163635,
+      "learning_rate": 5.681595362404312e-07,
+      "loss": 0.1107,
+      "step": 18360
+    },
+    {
+      "epoch": 480.26143790849676,
+      "grad_norm": 0.3596895635128021,
+      "learning_rate": 5.506948171566273e-07,
+      "loss": 0.1122,
+      "step": 18370
+    },
+    {
+      "epoch": 480.52287581699346,
+      "grad_norm": 0.3319242298603058,
+      "learning_rate": 5.335019886757442e-07,
+      "loss": 0.1116,
+      "step": 18380
+    },
+    {
+      "epoch": 480.7843137254902,
+      "grad_norm": 0.34595364332199097,
+      "learning_rate": 5.165810978023044e-07,
+      "loss": 0.1109,
+      "step": 18390
+    },
+    {
+      "epoch": 481.0457516339869,
+      "grad_norm": 0.3366650640964508,
+      "learning_rate": 4.999321907973698e-07,
+      "loss": 0.1103,
+      "step": 18400
+    },
+    {
+      "epoch": 481.3071895424837,
+      "grad_norm": 0.31628909707069397,
+      "learning_rate": 4.835553131784298e-07,
+      "loss": 0.1128,
+      "step": 18410
+    },
+    {
+      "epoch": 481.5686274509804,
+      "grad_norm": 0.33149656653404236,
+      "learning_rate": 4.6745050971923607e-07,
+      "loss": 0.1119,
+      "step": 18420
+    },
+    {
+      "epoch": 481.83006535947715,
+      "grad_norm": 0.30861008167266846,
+      "learning_rate": 4.5161782444971267e-07,
+      "loss": 0.1106,
+      "step": 18430
+    },
+    {
+      "epoch": 482.09150326797385,
+      "grad_norm": 0.357360303401947,
+      "learning_rate": 4.360573006558122e-07,
+      "loss": 0.1112,
+      "step": 18440
+    },
+    {
+      "epoch": 482.3529411764706,
+      "grad_norm": 0.3232167065143585,
+      "learning_rate": 4.207689808794046e-07,
+      "loss": 0.1088,
+      "step": 18450
+    },
+    {
+      "epoch": 482.6143790849673,
+      "grad_norm": 0.31159424781799316,
+      "learning_rate": 4.0575290691817757e-07,
+      "loss": 0.1121,
+      "step": 18460
+    },
+    {
+      "epoch": 482.87581699346407,
+      "grad_norm": 0.4624285101890564,
+      "learning_rate": 3.91009119825525e-07,
+      "loss": 0.1121,
+      "step": 18470
+    },
+    {
+      "epoch": 483.1372549019608,
+      "grad_norm": 0.34504663944244385,
+      "learning_rate": 3.765376599103587e-07,
+      "loss": 0.1135,
+      "step": 18480
+    },
+    {
+      "epoch": 483.39869281045753,
+      "grad_norm": 0.3002839982509613,
+      "learning_rate": 3.623385667371304e-07,
+      "loss": 0.1101,
+      "step": 18490
+    },
+    {
+      "epoch": 483.66013071895424,
+      "grad_norm": 0.37612277269363403,
+      "learning_rate": 3.484118791255986e-07,
+      "loss": 0.112,
+      "step": 18500
+    },
+    {
+      "epoch": 483.921568627451,
+      "grad_norm": 0.4303232431411743,
+      "learning_rate": 3.347576351508064e-07,
+      "loss": 0.1127,
+      "step": 18510
+    },
+    {
+      "epoch": 484.1830065359477,
+      "grad_norm": 0.4048657715320587,
+      "learning_rate": 3.2137587214293717e-07,
+      "loss": 0.1125,
+      "step": 18520
+    },
+    {
+      "epoch": 484.44444444444446,
+      "grad_norm": 0.2931053638458252,
+      "learning_rate": 3.0826662668720364e-07,
+      "loss": 0.1138,
+      "step": 18530
+    },
+    {
+      "epoch": 484.70588235294116,
+      "grad_norm": 0.29800963401794434,
+      "learning_rate": 2.954299346238032e-07,
+      "loss": 0.1096,
+      "step": 18540
+    },
+    {
+      "epoch": 484.9673202614379,
+      "grad_norm": 0.34082716703414917,
+      "learning_rate": 2.828658310477406e-07,
+      "loss": 0.1086,
+      "step": 18550
+    },
+    {
+      "epoch": 485.2287581699346,
+      "grad_norm": 0.38244274258613586,
+      "learning_rate": 2.705743503088165e-07,
+      "loss": 0.1134,
+      "step": 18560
+    },
+    {
+      "epoch": 485.4901960784314,
+      "grad_norm": 0.3117094039916992,
+      "learning_rate": 2.585555260114614e-07,
+      "loss": 0.1078,
+      "step": 18570
+    },
+    {
+      "epoch": 485.7516339869281,
+      "grad_norm": 0.32655006647109985,
+      "learning_rate": 2.468093910146685e-07,
+      "loss": 0.1124,
+      "step": 18580
+    },
+    {
+      "epoch": 486.01307189542484,
+      "grad_norm": 0.4254554510116577,
+      "learning_rate": 2.3533597743194967e-07,
+      "loss": 0.1127,
+      "step": 18590
+    },
+    {
+      "epoch": 486.27450980392155,
+      "grad_norm": 0.3003155291080475,
+      "learning_rate": 2.2413531663115773e-07,
+      "loss": 0.1097,
+      "step": 18600
+    },
+    {
+      "epoch": 486.5359477124183,
+      "grad_norm": 0.3139316439628601,
+      "learning_rate": 2.1320743923447517e-07,
+      "loss": 0.1114,
+      "step": 18610
+    },
+    {
+      "epoch": 486.797385620915,
+      "grad_norm": 0.38288217782974243,
+      "learning_rate": 2.0255237511830338e-07,
+      "loss": 0.1145,
+      "step": 18620
+    },
+    {
+      "epoch": 487.05882352941177,
+      "grad_norm": 0.3423730134963989,
+      "learning_rate": 1.921701534131848e-07,
+      "loss": 0.1115,
+      "step": 18630
+    },
+    {
+      "epoch": 487.32026143790847,
+      "grad_norm": 0.40306052565574646,
+      "learning_rate": 1.8206080250372515e-07,
+      "loss": 0.1129,
+      "step": 18640
+    },
+    {
+      "epoch": 487.58169934640523,
+      "grad_norm": 0.30820974707603455,
+      "learning_rate": 1.7222435002847147e-07,
+      "loss": 0.1108,
+      "step": 18650
+    },
+    {
+      "epoch": 487.84313725490193,
+      "grad_norm": 0.3316936790943146,
+      "learning_rate": 1.6266082287994533e-07,
+      "loss": 0.1087,
+      "step": 18660
+    },
+    {
+      "epoch": 488.1045751633987,
+      "grad_norm": 0.32804399728775024,
+      "learning_rate": 1.5337024720445403e-07,
+      "loss": 0.1149,
+      "step": 18670
+    },
+    {
+      "epoch": 488.36601307189545,
+      "grad_norm": 0.36528879404067993,
+      "learning_rate": 1.443526484020574e-07,
+      "loss": 0.111,
+      "step": 18680
+    },
+    {
+      "epoch": 488.62745098039215,
+      "grad_norm": 0.336240291595459,
+      "learning_rate": 1.3560805112655673e-07,
+      "loss": 0.1123,
+      "step": 18690
+    },
+    {
+      "epoch": 488.8888888888889,
+      "grad_norm": 0.4585813283920288,
+      "learning_rate": 1.2713647928532802e-07,
+      "loss": 0.1106,
+      "step": 18700
+    },
+    {
+      "epoch": 489.1503267973856,
+      "grad_norm": 0.3581952452659607,
+      "learning_rate": 1.1893795603932222e-07,
+      "loss": 0.1135,
+      "step": 18710
+    },
+    {
+      "epoch": 489.4117647058824,
+      "grad_norm": 0.3902879059314728,
+      "learning_rate": 1.1101250380300965e-07,
+      "loss": 0.1083,
+      "step": 18720
+    },
+    {
+      "epoch": 489.6732026143791,
+      "grad_norm": 0.2964220345020294,
+      "learning_rate": 1.0336014424424668e-07,
+      "loss": 0.1116,
+      "step": 18730
+    },
+    {
+      "epoch": 489.93464052287584,
+      "grad_norm": 0.39717039465904236,
+      "learning_rate": 9.598089828430911e-08,
+      "loss": 0.1142,
+      "step": 18740
+    },
+    {
+      "epoch": 490.19607843137254,
+      "grad_norm": 0.3992186486721039,
+      "learning_rate": 8.887478609777011e-08,
+      "loss": 0.112,
+      "step": 18750
+    },
+    {
+      "epoch": 490.4575163398693,
+      "grad_norm": 0.34660351276397705,
+      "learning_rate": 8.204182711246677e-08,
+      "loss": 0.1128,
+      "step": 18760
+    },
+    {
+      "epoch": 490.718954248366,
+      "grad_norm": 0.29538923501968384,
+      "learning_rate": 7.54820400094558e-08,
+      "loss": 0.1075,
+      "step": 18770
+    },
+    {
+      "epoch": 490.98039215686276,
+      "grad_norm": 0.43791237473487854,
+      "learning_rate": 6.919544272293577e-08,
+      "loss": 0.1137,
+      "step": 18780
+    },
+    {
+      "epoch": 491.24183006535947,
+      "grad_norm": 0.35819703340530396,
+      "learning_rate": 6.318205244023601e-08,
+      "loss": 0.1113,
+      "step": 18790
+    },
+    {
+      "epoch": 491.5032679738562,
+      "grad_norm": 0.31988316774368286,
+      "learning_rate": 5.7441885601716707e-08,
+      "loss": 0.1108,
+      "step": 18800
+    },
+    {
+      "epoch": 491.7647058823529,
+      "grad_norm": 0.3159491717815399,
+      "learning_rate": 5.19749579007911e-08,
+      "loss": 0.114,
+      "step": 18810
+    },
+    {
+      "epoch": 492.0261437908497,
+      "grad_norm": 0.3267784118652344,
+      "learning_rate": 4.678128428382555e-08,
+      "loss": 0.1088,
+      "step": 18820
+    },
+    {
+      "epoch": 492.2875816993464,
+      "grad_norm": 0.3536956012248993,
+      "learning_rate": 4.186087895011737e-08,
+      "loss": 0.1136,
+      "step": 18830
+    },
+    {
+      "epoch": 492.54901960784315,
+      "grad_norm": 0.3293076455593109,
+      "learning_rate": 3.721375535188365e-08,
+      "loss": 0.1121,
+      "step": 18840
+    },
+    {
+      "epoch": 492.81045751633985,
+      "grad_norm": 0.2942824065685272,
+      "learning_rate": 3.283992619416143e-08,
+      "loss": 0.1117,
+      "step": 18850
+    },
+    {
+      "epoch": 493.0718954248366,
+      "grad_norm": 0.34250175952911377,
+      "learning_rate": 2.873940343485204e-08,
+      "loss": 0.1103,
+      "step": 18860
+    },
+    {
+      "epoch": 493.3333333333333,
+      "grad_norm": 0.3269750475883484,
+      "learning_rate": 2.4912198284621214e-08,
+      "loss": 0.11,
+      "step": 18870
+    },
+    {
+      "epoch": 493.5947712418301,
+      "grad_norm": 0.339807391166687,
+      "learning_rate": 2.135832120689907e-08,
+      "loss": 0.1081,
+      "step": 18880
+    },
+    {
+      "epoch": 493.8562091503268,
+      "grad_norm": 0.32826676964759827,
+      "learning_rate": 1.8077781917846815e-08,
+      "loss": 0.1127,
+      "step": 18890
+    },
+    {
+      "epoch": 494.11764705882354,
+      "grad_norm": 0.29220259189605713,
+      "learning_rate": 1.5070589386345645e-08,
+      "loss": 0.1146,
+      "step": 18900
+    },
+    {
+      "epoch": 494.37908496732024,
+      "grad_norm": 0.3818574547767639,
+      "learning_rate": 1.2336751833941229e-08,
+      "loss": 0.1145,
+      "step": 18910
+    },
+    {
+      "epoch": 494.640522875817,
+      "grad_norm": 0.3570195138454437,
+      "learning_rate": 9.876276734832601e-09,
+      "loss": 0.1106,
+      "step": 18920
+    },
+    {
+      "epoch": 494.9019607843137,
+      "grad_norm": 0.2886011600494385,
+      "learning_rate": 7.689170815872172e-09,
+      "loss": 0.1096,
+      "step": 18930
+    },
+    {
+      "epoch": 495.16339869281046,
+      "grad_norm": 0.36767634749412537,
+      "learning_rate": 5.775440056521308e-09,
+      "loss": 0.1102,
+      "step": 18940
+    },
+    {
+      "epoch": 495.42483660130716,
+      "grad_norm": 0.30554234981536865,
+      "learning_rate": 4.1350896888503464e-09,
+      "loss": 0.112,
+      "step": 18950
+    },
+    {
+      "epoch": 495.6862745098039,
+      "grad_norm": 0.35083529353141785,
+      "learning_rate": 2.768124197505273e-09,
+      "loss": 0.1123,
+      "step": 18960
+    },
+    {
+      "epoch": 495.9477124183006,
+      "grad_norm": 0.37509220838546753,
+      "learning_rate": 1.6745473197188333e-09,
+      "loss": 0.1122,
+      "step": 18970
+    },
+    {
+      "epoch": 496.2091503267974,
+      "grad_norm": 0.3241898715496063,
+      "learning_rate": 8.543620453105306e-10,
+      "loss": 0.1103,
+      "step": 18980
+    },
+    {
+      "epoch": 496.47058823529414,
+      "grad_norm": 0.40990370512008667,
+      "learning_rate": 3.075706166089098e-10,
+      "loss": 0.1103,
+      "step": 18990
+    },
+    {
+      "epoch": 496.73202614379085,
+      "grad_norm": 0.401498407125473,
+      "learning_rate": 3.417452852927383e-11,
+      "loss": 0.1131,
+      "step": 19000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 19000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 500,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3677835109083546e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}